[Open Source Publishers] Rudder: cybersecurity tool, self-financed company for 15 years

Rudder, cybersecurity tool, self-financed company, with Alexandre Brianceau

Walid: Hello and welcome to Projets Libres, LinuxFr.org’s podcast that talks about free software, open data and digital commons. I am Walid Nouh and today with me, I have Alexandre Brianceau who is the CEO of Rudder. Rudder is a French tool for infrastructure automation and compliance. You’ll see, we’re both going to talk about what Rudder is, how it came about, where it goes, its business model. It’s going to be really interesting.

Alexandre, listen, I’m really excited to have you on the podcast. We’ve been trying to record for a long time. So there we are, here we are. So, welcome to you.

Alexandre: Thank you Walid. and then hello and thank you for welcoming me and it’s true that we’ve been discussing it for a while and finally it’s projecting so it’s cool I’m happy And without a megaphone, unlike usual, on OSXP [Open Source Experience].

Presentation of Alexandre Brianceau

Walid: That’s it, that’s it. Listen, we’re used to it, we’ve already crossed paths quite a few times in the living rooms, but I’ll let you introduce yourself for the listeners of the podcast who don’t know you yet.

Alexandre: I’m Alexandre Brianceau, I work for Rudder, I became its CEO. So, I didn’t start like that. I started as an employee a little over eight years ago now: project management, project management, and then I climbed the ladder a little bit by little. We will talk about it later in the genesis and continuity of the product.

I’ve always worked in the free software industry. I did university studies, first on infrastructure, a bachelor’s degree, and then on social security for my master’s degree and always around free software. And I joined Bull at the time in my early years and I did a lot of free software and I did security, in particular BCP/DRP, so disaster recovery. It was the beginning of the cloud, all that at the time it was in the 2010s and I worked a lot on Red Hat technologies.

So I did a lot of that, then I went to live in French Polynesia where I worked for the Polynesian government’s infrastructure, always with a lot of free software. I met in France, I worked for Worteks for a few months to help bootstrap the company and then I went on to join Normation, which is Rudder today. Same, we’ll talk about it a little later. So I have a rather tech profile, very free software oriented.

Walid: Okay, super interesting. We’ll talk about it later on what it is to have a tech profile and after having to manage a company and a software publisher. I think it’s not the same, so it’s going to be quite interesting. Old-fashioned bull!

Alexandre: Yes, definitely. There are still many of us who recognize each other. It was a big family.

Rudder Overview

Walid: So, let’s get to the heart of the matter. The first thing I’m going to ask you is, I think there are quite a few listeners who don’t know Rudder. So, can you explain what Rudder is, please?

Alexandre: So Rudder, in broad strokes: we are a French software publisher. So we’re developers, we do consulting and training, but it’s really associated with Rudder only, with the product. Our goal is to help ops to be able to secure their infrastructure, quite simply, and therefore to automate it. I come from the social security sector, the software came from the management of S-Code infrastructure accounts, so we’ll talk about it in the genesis more entirely, but our mission is really to help automate the management of Linux and Windows infrastructure, and in particular for the Linux podcast part since it’s on the Projets Libres podcast. That’s our only activity today: to develop and implement Rudder for our customers, or sometimes for our users who also become customers because they will pay for services from us to go and deploy there.

Today, we are a company of 24 people. It’s evolved quite a bit. There, I was just doing the launch of the year last week of the company. In three years, we have achieved 60% HR growth. We went from a very small business of about ten people to, so today, 24 people. It changes everything about the box, its structure, what comes next. And so there you have it, it’s interesting. Then the company is 15 years old. We celebrated the 15th anniversary of the company last year.

For 12-13 years, it has remained quite stable, and now we are growing so it’s cool, and at the same time pass your new challenges if not to talk maybe more about the product as such, since your question was also focused on the box because I also think that the product since we have the same name well the same original site is called Normation. There may have been a lot of people who knew the founders and our solution through the name Normation was our original name, but as it turns out that now we only do Rudder development, we are called Rudder. So Rudder, you presented it very well, it’s an automation tool, a platform: I call it cyber automation for Ops. We rely concretely on infra-as-code to help Ops to be able to make reliable, harden, update: so we do patch management, compliance with security standards, and therefore infra-as-code.

It looks like Puppet, Saltstack, except that we have a French, European solution. It’s quite original because the social security tools usually target social security people. We come from the Ops world and the objective was to bring security to Ops with a tool that is designed for production, which is not necessarily very common. And we add a little extra difficulty because it’s a bit of a particular market and we’re not so much in hype when we talk about toughening, updating, complying with a security policy is not super sexy, but for all that, it’s the daily life of many people today.

So we decided to help these people in the long term with a solution in mind, a source precisely for this sustainable and resilient side as well. So our main features are patch management, continuous deployment, maintenance, so the idea of idempotency: we act permanently on the systems, in fact every 5 minutes by default, we come to make sure that it complies or not with the state of the art what we want on our infrastructure with a centralized infrastructure and with a no-code approach.

So it’s infra-as-code with a no-code part, there are both with a YAML part and a visual part, which makes the tool also accessible beyond sysadmin. This is also often the problem that has been reported to us: infra-as-code tools have been around for a long time, there are a handful of experts in the companies who master the tool: it’s infra but it’s so critical that there is visibility to be given to management or its customers or to the security teams and therefore us, In the genesis of the product, we arrived right at the time of the creation of DevOps the beginning of the advent of the cloud and so it made sense to have a tool that responds to the challenges of the DevOps culture, to bring visibility too, not just tools and automation, to know why we do it, so there you have it, it’s a simple and accessible product that doesn’t remain too powerful.

Walid: I discovered it at the time of the launch because it was quite related to what we were doing more on fleet management, I even think that I must have done a CFEngine training a long time ago with you.

Alexandre: Yes, we started like that, absolutely.

Rudder’s license

Walid: What is Rudder’s license?

Alexandre: It’s LPG V3. This has not always been the case. We can talk about it here or later.

Walid: We’ll talk about it later when the thing comes about.

Alexandre: It’s GPL V3, which means that we have a dual licensing, especially for those who want to do business with the product, and at the same time it’s a real open source license and gives users full use of the product. With an Open-core perspective, which I think we will talk about later on the Business Model. But here’s the thing, it’s not fictional open-core in the sense of the demo or anything.

We have segmented our product: the main platform is the same for everyone, customer or Rudder user, to do security-oriented configuration management. We just add more social security and enterprise-oriented features to the paid offer.

Walid: To learn more about open core and certainly other terms that we are going to talk about, I invite you to refer to the episode that I will of course put in the transcript, on the introduction to the economic models of free software that we did some time ago with my colleagues Raphaël Semeteys and Goneri Le Bouder. That’s it, and so everything is explained in there and certainly in many other episodes.

Alexandre: It’s a big subject because in addition, there are a lot of changes over time, especially in recent years: changes in licenses or business models. It’s interesting to know where to put your feet and what I also like is that in the podcast it takes stock of the different models without saying that this or that model is the first to take or the main one to take, even if we see that there are SaaS trends with open source software and financing through SaaS. But that’s not the only way to prove it, we’ve been growing for a few years with the old Open-core model, but it still works very well, there are some that are Zabbix , 100% open source and that work great too.

The technologies on which Rudder is based

Walid: What are the technologies on which Rudder is based? What language do you use? Which technology? Big tech, in other words.

Alexandre: In terms of language, we chose functional languages because already, they are languages that are more likely to work with small teams and we are not intended to be a big publisher. And then, it makes it easier to do refactoring. There are a lot of interesting things with functional language. It’s okay with Rust, on Scala, there, do Elm, do the graphic part. And yes, that’s cool.

And on top of that, it allows us to have people who are very interesting in these communities. We really have no trouble recruiting. When I talk to other devs who are on more mainstream technologies, like PHP or others, Python, etc., it’s not always easy. There is a little to eat and drink. Here, it’s communities that are tighter, so in the end, they are hyper relevant for us, both to learn and to recruit, or even to watch people too, because there aren’t 36,000 companies doing Scala in France, for example. That’s cool.

And then also we have a lot of libraries and then software that we also use that are included in the product, maybe the one that is the most famous, let’s say will be Fusion Inventory , I think that by the way, you have a bit of a privileged link with this project.

Walid: I have a certain history with this project indeed [Walid was part of the project for several years].

Alexandre: And so we’re super happy and that’s it and it works, so it’s the principle of free software, so we use and contribute, either it’s through code or it’s going to be financially in some cases.

It’s the same type that, if we get out of technologies that are open source communities too, like April or Framasoft , the whole Framasoft team is important things, there are plenty of ways to contribute to free software. It’s always worth remembering it in each episode, it’s not just the financial, it’s not just the code, there are many other things, even if it’s just talking about the software around you, it’s very nice.

Walid: I’m trying to remember but I think you were the first company I heard of at the time that made Scala, at least in my entourage…

Alexandre: And Rust is the same, now it’s still a fairly well-known language and it works better, well a little more known than Scala, but it’s still not confidential, which is still more restricted than Go or others.

But here we are, we have an extra interest in it because we make a lot of systems, we do a lot of dev systems, so we need to have a language that is able to be adapted to the system orientation and that is more modern than C.

Walid: or that the Perl.

Alexandre: that Perl yes good kind I don’t even talk about it anymore.

Rudder’s competitors

Walid: The next question to make the panorama of Rudder is for people who don’t know Rudder, do you have any competitors? And if so, can you name a few of them so that we can put it in perspective in relation to its ecosystem?

Alexandre: Absolutely. So now, it’s worth it, maybe I’ll reintroduce the fact of having the Open-Core part a little bit.

So the Open Source part is free to use, it’s called RudderCore. It does security-oriented configuration management: it has competitors such as Puppet and SaltStack. The most common is Puppet which has not sourced this part too much, it is rather in the paid parts, whether it is the graphical interface or the slightly advanced features.

And on this point, in fact, we make available free of charge, in addition to being free, a very large part of what it is. We have some features, more on Enterprise, that are open source, but not made available through free binaries. By the way, which means that we have a lot of universities or agglomerations that use us in open source and that benefit from these Enterprise plugins, with the time it takes them to build each time. And that’s part of the goal of the game. I think we’ll talk about it in the business model.

There’s Ansible, which isn’t really a competitor quite, but which can have certain cases where functionally, depending on the needs of the teams, it can really overlap with teams that are a little bigger and a little more mature. We really see the distinction between deployment and maintenance. But it depends on the person. That’s to situate RudderCore.

And then, otherwise, we do patch management and compliance with security standards and vulnerability detection. So here, we are much closer to the more traditional security tools, always with a view to making it available to admins. That’s our added value in it and so it will put us more in confrontation with Qualys, Rapid7 which are not open source especially sometimes we are compared a little with Wazuh but the same I find that it’s more complementary than really competitive in our case here it allows us to extinguish I think a global panorama.

Walid: And to finish on this part of the presentation, are there regions where the Rudder is more popular than others where it is more popular, where there is perhaps more community?

Alexandre: Of course in France because we are French. So inevitably we come across trade fairs in France, and we have a network in France.

Now Rudder, we have chosen to develop it in English without the possibility of translation: it’s rather not a default choice at that time but here is the doc, everything is in English, the name is in English too in order not to make it finally not to make a Franco-French open source project. Which means that in fact it’s not particularly attached to France in terms of development, even in terms of turnover for us.

Alexandre Brianceau

In France it works quite well and in Europe, in general, it works quite well, in particular, as I said, most of our tools that I mentioned are rather American or Israeli tools. And so, in the current dynamics of resilience, of the desire to repatriate from there, the fact of having open source solutions is a first thing. And by the way, there is the DINUM, we are working very actively on this. She is not the only one. There are a lot of French open source companies that work very well for these reasons. And so in Europe there is still a revival but we also find it in other countries, even Anglo-Saxon countries, but which would be in a desire to separate a little from the Americans, so I am thinking in particular of Canada, Australia which is starting to work well for us too, which means that today the sun never sets on Rudder somewhere but which is rather recent for us we have never tried too hard to develop it particularly.

Walid: Yes, that means that afterwards, you have to do 24-24 support with people maybe in different time zones and everything. That’s something else [laughs].

Alexandre: We manage it with a specific dynamic. It’s because, as I said, we’re publishers. So, in fact, they don’t want to create an arm’s length relationship with us. So our goal is really to train people as much as possible and not to sell days. Finally, I don’t need to sell consultants. It’s not my business model at all. And then, it’s a bit part of the free spirit also for us, knowledge is made to be shared, so I actually prefer 24-7 support. Well, we made a bit of extensive support because accountants in some big companies, I was thinking in particular in Canada, we have the equivalent of the local Française des Jeux: well, they themselves, for legal principles, they need certain guarantees, we did it a little for them, but in fact they almost don’t use it, we manage to work well differently.

So the open source mindset, including to do business outside of making open source software, but it’s a win-win, it’s transparent and in fact we are in a relationship of trust with our customers and that changes the dynamic a little bit, necessarily for the positive.

Rudder’s Creation

Walid: Now let’s talk about the genesis of Rudder, I knew Rudder at the very beginning, so I would like you to explain…

Alexandre: I also knew Rudder on Solutions Linux in 2012 or something like that, it was a long time ago.

Walid: The great era so you started to touch on two words but can you go into a little more detail on the questions, the reasons that led to the development of Rudder?

Alexandre: yes definitely.

Walid: At the time, what was the panorama and why did we feel the need…

Alexandre: I can’t tell you that because the company is above all a human adventure also at the base so I can’t tell you that without saying that the three founders actually worked together or were led to work together and so they were already in free software and either in consulting or in rather system dev but suddenly for critical infrastructure.

They found that it was repeated over and over again. They were always trying to autodomatize things to allow sysadmins to take time for themselves and do more added value than repeating the same thing 100 times: filling Excel tables right, left, changing a few lines in confs on 20, 30, 50, 100 machines. And suddenly, they had the idea of developing a product. So, there were already a few products that existed to make infrastructure as-code. At the time, there was CFengine, there was Chef who was starting to take his nose. Very quickly, afterwards, behind the Puppet, there was Ansible, then Saltstack which arrived but so they developed their they created their company for that with suddenly as an overview the beginning of the advent of the cloud as I said so we were starting to enter precisely into the pet versus cattle so the objective to be able to say I manage more I machine by machine but rather by group of machines and it was the beginning of DevOps as well and that’s when we realized that the as-code infrastructure tools that we had to deploy or that we were trying to administer or deploy for customers had been able to use it. And so, we developed ours for that reason.

And so, Rudder’s goal, from the start, was to help ops save time and simplify their infrastructure management by making the infrastructure as-code accessible and gaining visibility. So, as I said, human adventure also above all, because the same, I can’t not link it, because it’s still part of our main driver today in terms of gearbox control, it’s not to do like the classic ESNs — let’s say those who work in the infra or in tech in general we’ve gone through the ESNs or almost where we know people who are in it — we didn’t find ourselves so the founders at the base and I don’t find myself in the values that are carried by this type of company either, so looking for something more sustainable, to be the company that we would like to have as an employee and that necessarily involves the creation of a company and not just the creation of a free software project to do in your free time or to to be financed by the time you spend as an employee because there are plenty of people who create open source products, especially OVH , they are still quite specialized in that, for example, it’s cool, but at the same time there was this dimension of creating a kind of community of companies that are

Walid: the people who are behind it?

Alexandre: So there’s Jonathan Clark, François Armand, Nicolas Charles , that’s the three names, I think they’ll speak to you if you’ve known the company for a long time and who are still in the company today. Now, we have quite different roles. François became the CTO. He was already quite good in there, but he stayed. It’s the only one that really stayed in a direction really displayed in the box. He and I, now, manage the company together, mainly. Let’s say, on a daily basis, in a daily direction.

And then there’s Nicolas who takes care of all the Customer Services part, the deliveries, the use of prowlers, the community, so which is very important too. And John, who takes care of the administrative and financial part, which is also the backbone, without that, we can’t do anything in the company, they’re still there, and that’s it, it’s just that now I’ve become the boss, but somehow I’m still working for them since they remain the majority shareholders of the company.

Walid: hello to them, we always meet at trade fairs, so hello to them.

Alexandre: It was very happy to know that we were going to talk together.

Walid: but yes every time we cross each other in fact our paths crossed quite a few times at that time and from the beginning that’s why it’s so nice it reminds me a lot of things from the time. We also tried to do compliance stuff at the time with GLPI, Fusion Inventory and everything, but there wasn’t really any infra-ascode yet. We had never managed to do that, and anyway, we were much more on the post-user part than the server part at the time.

Alexandre: Yes, completely, but it makes sense, in fact, it’s a little different needs. And it’s good that I’m telling you, it’s not easy. In fact, I know that there are plenty of people who think it’s easy to do. You have to remember that most of the big competitors I mentioned are hundreds of people who work for these companies. This is no longer the case for Puppet and Saltstack which are completely collected, but they are still very big investments. And managing systems in an automated way, we still often realize this from customers or potential users who talk to me about making scripts. In particular, for example, for patch management, which is a little more recent in our history. I have scripts to make updated deployments. Yes, of course, it can work, but in fact, it’s much more complicated than that if you ever want to scale on hundreds or thousands of machines, for those who are behind the scenes, they know that very quickly you find yourself with big problems, things that don’t scale or don’t work the way you want and you, as a software publisher, can’t Allowing yourself to have a wobbly thing is not a piton script quickly deposited on a GitHub and everything is reused.

Rudder’s starting business model

Walid: We introduced Rudder quite a bit, let’s talk about economic models at the time, by the way, when the project is created, it’s a project, a company, so a publisher at the base, what is the starting model? How did you make money back then?

Alexandre: The starting point, before the product existed, because it took a little bit of development time before it existed, was consulting. That’s what made it possible to bootstrap the box. By the way, I can also get the point out of the way quite quickly, we have been self-financed since the beginning and this is still the case today.

Walid: We’ll talk more about it after that.

Alexandre: So,

This explains why consulting is that it’s a way to be able to start financing the time we have in intercontract, let’s say, in quotes to develop the software so there was that and in fact afterwards very quickly it became the support of free software which worked well at the time it was done a lot it is still done but it was almost the majority at the time in the years 2010 after that, there was also sponsored development because the first users were advising on the deployment, training and these rats, I’d like that or that, so we offered them to develop features in a sponsored way, so they paid us dev time to develop the feature. So that was also the case for quite a while.

Alexandre Brianceau

So it was the economic model when he left that has evolved a lot since then. I don’t know if you want me to go on a little bit about that too.

Walid: Before moving on to this, at the time, when you created the project, you had chosen which license and why?

Alexandre: The AGPL.

Walid: AGPL?

Alexandre: yes, the AGPL, because precisely, the objective was precisely to be able to perpetuate the fact that we could make a business model other than just support and in fact we had to change that. So the objective was to make people quite free, but at the same time to own your code so that you don’t have everything you need, all the cards in your hands to avoid getting stuck with doors, later in the rest of the product, we didn’t really know, well, they didn’t really know at the time where we were going, I don’t know in 10 years what it will be like either But in any case, they made choices that were relevant because in the end today we still own our code even if it’s free and we do what we want with it, which, like self-financing, is very interesting when we have a box and company sustainability dimension that is very important to us.

The AGPL allowed us to protect ourselves. You also have to look at the fact that among our first users, we have a lot of outsourcers. When we talk about administering systems automatically, especially in security, a lot of outsourcers. It’s to make sure that one day, an outsourcing company will come and monetize the use of Rudder without paying us a part, without retro-commissioning. Today, it’s becoming a daily topic: GAFAM are sucking up open source projects or companies in general. It’s an Arlesian one between free software players, we still talk about it often. And so, the AGPL has the benefit in theory of being able to help with this, even if it remains, you have to be able to detect it, prove it and all that. That is another subject.

And we have changed on that. Why? Because in fact, we had our first big industrial customer in the automotive industry, whose lawyers didn’t understand the AGPL and it even scared them, these restrictions. As a result, we had to switch, we were able to switch to the GPL V3. It happened quite quickly in the first years of the company: it wasn’t too much of a subject since in the end it granted more rights to users so it was well received where, on the other hand, going from a more open license to a GPL can cause problems there, it was less the case.

Alexandre Brianceau

Walid: You didn’t specify it but so it means that any outsider who contributes to Rudder signs a CLA agreement?

Alexandre: Exactly, absolutely. In any case, for those who want to use it and develop or contribute in one way or another, the code is actually modified. So, we still regularly have people who sign us. I would say, on a monthly basis, we have new users who make themselves known through this means. Even if they are people with whom we can’t exchange. We’ll talk about it later when we talk about the community, but it’s complicated for us to exchange with the community. OK.

Walid: Just, if it is a Contributor License Agreement, to find out more, go see the episode ongovernance models with Sébastien Dinot. He talks about all this, so everything is explained very well.

Alexandre: Yes, yes, absolutely. But I think there are a lot of people who do it and never sign a CLA. This is not serious in itself. In any case, it’s still everyone’s own will. There are broad outlines and in any case, it’s made to be on the brake in one way or another. The key is to have the means to be able to defend ourselves if one day it comes against us.

Walid: The important point in this is that you had an AGPL, but since you get an agreement signed, a CLA, then you have ownership of the code, so in fact, you didn’t have to ask each of the contributors if they agreed to change the license.

Alexander: Exactly, absolutely, but we have very few contributions back and especially here the companies for having discussed it with several users who have been users for a long time or even customers and in fact they have trouble getting their company to sign because suddenly there is an individual or a corporate, Well a company and at the company level it’s often badly perceived because in fact today the R&D development that they can do including on open source software, they manage it more often with IT and they protect themselves Especially since it’s often industrialists who are able to do it with a big telecom connotation. I’m talking about this because it’s a French-language podcast. Telecoms is our great history in France. But the fact remains that IT protections are very structuring in the economic construction of France around this telecom subject. And so, it remains a major obstacle to being a center and signing companies.

Walid: Ok, so now you’re going from AGPL to LPG.

Alexandre: Afterwards, we stayed with LPG V3, but we also started to develop parts that are proprietary and it came later in the company. So this is not a change of license because they were new products. I hope that in the long run, we will be able to earn far too much money for what we need. That’s kind of the goal. Today, for example, we don’t pay ourselves any dividends to shareholders and there are no plans to do so. And so, the goal is that one day, we will make enough money so that we can opensource what we do, or at least, a large part of what we do, because we don’t take the rug from under our feet.

Walid: And there, typically, you talked about dual licensing. Did that appear even later? Or does it appear at the same time as the passage?

Alexandre: At the same time. And in fact, finally, today, we have technological partnerships that can happen, but in fact, it’s a discussion. It’s just that it necessarily leads to a discussion between the two entities and which is a good thing to do. Does it allow you to build something instead of collaborating together on the construction rather than doing it on someone’s back?

Walid: Can you explain how it works, precisely, for people who don’t stand for re-election, the dual license, what it’s for, and how it happens when someone comes to you and you direct them to the proprietary license?

Alexandre: Let’s say that,

For example, a GPL, if you modify or make changes to your code, you are obliged to be able to retro-contribute it and therefore give us ownership of what you do, which is not necessarily the wish of all companies, and in particular if you ever want to integrate Ruddor as a white label, for example if you are an outsourcer, you want to integrate it, make custom developments for your own needs And having a competitive advantage over your outsourcing competitors, well why not it’s even a good idea rather than reinventing the wheel but for all that you don’t want to give it back to the publisher behind it either, even if they like us, even for people we’ve known for a long time, it’s not really what they want to do. So we discuss it together and we see what we can do and in fact we can give them a license which is it’s the same code it’s the same thing except that they are not obliged to retro-contribute and we have made our arrangement between us which can be financial or other it is not necessarily financial the relationship between the two

Alexandre Brianceau

Walid: When you do this kind of thing, is it just a purely contractual thing or do you need lawyers?

Alexandre: So, the contract, I always advise you to the lawyers, as a general rule. I have a lot of them, depending on what subject it is. But yes, it’s here, it’s contractual, indeed.

What I say every time about contracts is that as long as everything is going well, everything is fine. The problem is when things don’t work anymore. It must be foreseen and anticipated. If what we’re doing, in terms of software development, has the potential to be adapted by our users or our customers, to what extent do we want it to contribute to a community, and that can be the case? For us, it wasn’t necessarily the case.

We made the software open source so that it could be easily readable and modified by people, but not to create a community of creators and developers. While there is a lot of free software, it’s the opposite. They will also look for a sum of contributors through open source that they would not otherwise have. And so, in our case, it was worth protecting and anticipating it in this way. Others who are looking for a hyper-community development will turn to much more free licenses such as Apache or other types of licenses.

Alexandre Brianceau

The fact that Rudder is not open source

Walid: By the way, I didn’t note it in my questions but I think that in a discussion I don’t know if it’s with John or Nicolas who actually told me in real life if you go to the site and when we discuss we don’t necessarily highlight that the software is open source because that’s not necessarily what customers are looking for.

Alexandre: It’s part of the orientation of the business model and it’s debatable. For a very long time, we even had two different sites. It turns out that in fact, the subject of open source is no longer a subject in itself, it has become a strategic subject. When you search for a website, when you search for a project, either you know that you are looking for open source software and so basically, you go and look for GitHub and your type of software and that’s it. But otherwise, the first thing to do is that your software meets your business needs. And so, we respond to that. And so it’s a reflection on the user journey so we don’t really put it forward anymore even if in the menu it says community and very quickly you have an open source part to read to the GitHubs and all that but it’s also a desire to say in fact there are people who are a priori very fast that we won’t be able to remove open source equals free no well everyone knows it people who are very pro open source and who are informed about it those who are more open source users in the sense I rather want free software because I don’t have a budget because for many reasons we want to find ourselves making amalgams and what makes sales much more complicated there we talk more sales strategy after marketing strategy. That’s why we don’t put it forward anymore in particular, it’s also that today we want to highlight Rudder as a security automation platform and therefore the so-called Open Source, Rudder Core part of configuration management is now only one of our three solutions and so we were a little uncomfortable differentiating the two.

We found it more complicated so now we prefer to have a GitHub page dedicated to that and then the website which is rather oriented towards Enterprise and moreover whether it’s open source or not it’s not even the debate in terms of use or the way well there are people who need it to be open source to use it, that’s a first element but the fact that it’s open source, doesn’t make the software better or cheaper or whatever, besides, we have a very corporate positioning, Enterprise quite expensive, the fact that we’re open source doesn’t make the software cheap, it’s like buying organic, it’s not necessarily a reason to buy organic, maybe all products should be organic, maybe all software should be open source.

Rudder in SaaS?

Walid: So your current model is open core, you’re talking about a platform, so you have a SaaS platform too?

Alexandre: Actually Rudder is a software that is a tool that really integrates a lot of back-end business logic and that allows you to do a lot of different things afterwards. If I take, for example, the logic of group management or communication between the nodes and the central server, it’s common to all solutions, that’s why I talk about a platform in this context. And so in fact Rudder to do configuration management is the open source part that we call Rudder Core in fact on top of it we will add plugins.

We don’t put forward such marketing when talking about it because generally when we hear plugin we really expect just a small addition and not at all a real added value in the product so we present it as different solutions but in fact they are plugins in fact we thought of all the interfacing on purpose, we made the product modular so that the plugins can modify in depth, as if we were developing in Rudder’s heart. This allows us to have the same Rudder everywhere, for our Open Source users as well as for our customers. And by the way, our customers, there are some, they are customers of us, but they use the same Rudder as an Open Source user. Among the subscription we offer, we have private rest [depots], especially for longer-term maintenance, but there are people who don’t even use them, they are customers with us, but they continue to stay on public rests.

Walid: You said earlier that there were parts that were open source, but that were not packaged.

Alexandre: In user management, for example, LDAP Interconnection is really stupid and nasty.

And so, we took it more as how we do it so that a user, if he wants to pay us, and knows that his company can do it, how we do to give the marbles internally to demonstrate the value and buy the subscription. So there you have it, the Interco LDAP and Active Directory are open source, but we don’t make it available for free. That makes it a little easier to talk to the management on that side. That’s one example, but I’ll take you another one, for example, in the same genre. Linux agents are open source, but they are built and made available for free only for distributions that are currently maintained without scope. So typically, if you want Debian 9, you can’t get it for free. You have to build the agent yourself or you have a client with us, that’s the same, it works very well without you paying with us but it makes an easy argument to negotiate a subscription with your boss.

Alexandre Brianceau

Walid: I know we had the same problem at the time with inventory agents, if you wanted to have an agent on AIX or something else, you needed the subscription that gives access to the package. Which is understandable because it’s hyper-advanced, quite specific needs and everything.

Alexandre: It’s understandable, but it goes in all directions.

I mean, nothing would stop you from making it 100% open source and creating SaaS offerings and monetizing like that. As a company, in any case, we play for sustainability and we need to innovate. We need to innovate because the sector is constantly changing, because we are fighting against both open source and proprietary software, especially when it comes to security, this is always true. And so, to innovate, we need money. And there are plenty of ways to do that. This one is the one we have historically and that works well, so we are making it permanent.

Alexandre Brianceau

I remember that I had long debates, for example, on user rights for a very long time. If only to have user rights, so to say, so-and-so is in reader only, the other is in writing. It was open source, but not distributed for free. And now we’ve integrated it into the product. But it was part of the debates because, yes, it makes as much sense to put it on as it does not to put it on. And I completely understand a user point of view to say, it’s still part of the beaba to manage user rights in the product. And since it is forbidden to say “I need to have a kind of gatekeeper of incentives to pay for a subscription because even people with a better will sometimes can’t do it and that’s not counting all those who are with very bad will”.

Walid: in fact basically it’s from the moment you’re pretty clear on what you’re going to find in the packaged open source version and everything and what you’re going to find in the proprietary version and what you’re going to find if you make your own package yourself after I think it’s quite simple for users to understand the thing as long as it’s explained that’s often the problem What you have with the open core is that in fact what is in the open source version or the proprietary version is not necessarily very explained

Alexandre: It’s done on purpose, yes.

Walid: which means that in fact it creates unnecessary tensions and it creates unnecessary tensions, we didn’t explain to them that you will never find this thing in the open source version.

Alexandre: yes that’s right, you really explain open core, the bad open core is freemium In fact, it’s almost a proprietary software, but with a free part, you get your free dose. And then, once you’re captive, you’re forced to move on to a paid game.

There are also those that are not really open-core, but it is open-source software. But on the other hand, there are some parts that you can’t do without paying, such as updating your server easily.

Walid: Ah, Updates, migrations! It’s a classic trick.

Alexandre: Yes, exactly. You are forced to uninstall everything, to do a complete migration. Once again, I, now being on the other side, I take a step back by telling myself that he has to find a way to be able to pay himself but it creates brakes on use and so after all it’s to put the right cursor.

You can’t please everyone, I keep having people complain about saying Rudder is not really open source: yes I know and I’d like to do better, because I have 24 people to pay and I’d like to pay more because I think I’m doing a different company with our team and I’d like to continue to develop that. In short, we have strayed from the subject.

Walid: It’s a subject that comes up all the time in all the interviews we’ve done with people, with founders or managers of open source companies. It’s that at some point, you have to pay people and so at some point, you have to make compromises. And even if you’re convinced, at some point, you have to compromise somehow. By the way, you’ve never discussed making a SaaS platform yourself or was it a desire not to do it so as not to compete with people who would do it?

Alexandre: We’ve had competitors who have launched on SaaS platforms and have fallen apart. So, already, it didn’t make us want to do it. Secondly, creating a SaaS offer is still quite an investment, both in hardware [hardware], but that, I’m talking about it now with RAM prices at the moment, but also in time, and that also means, there, for 24-7, that we absolutely need, if you make a SaaS offer.

And Rudder has root and admin access everywhere on the infrastructure: we don’t want to bear the responsibility today, at least not without putting disproportionate resources into it to really make sure to our customers in complete transparency that we are confident on our platform. So that’s why we’ve never started with this model, we have regular discussions with outsourcers who are thinking of trying it or cloud providers who would like to be able to do this potentially. For the moment it hasn’t come to fruition but it could maybe succeed one day you never know, I’m not going to talk about it, it’s no longer at its stage. It’s not mature enough yet.

Setting up a self-financed company

Walid: We started talking about financing, but there is something very interesting. I’ll put a link to a conference you gave, I think, last year which was called… No, is Open Source dead?

Alexandre: Yes.

Walid: It immediately appealed to me when I watched the conference again, because you come back to the fact that you are self-financed. So, you don’t have an investor, you don’t have venture capital and everything. What does it change for you in terms of financing and also in terms of the freedom to manage your company? Can you tell us a little more about it? Because I find that really interesting. Every time we talk about this thing, with the different guests, we always come back to the subject of “Oh yes, but in fact, there are investors.” And so, if there are investors, overall, you’re not necessarily in control of what you’re going to do.

Alexandre: First of all, trigger warning, indeed we are pro self-financing. That doesn’t mean that we are against funders. And by the way, I have regular exchanges, especially with the people of Passbolt, for example, but they are not the only ones. Centreon there recently too. It’s a different dynamic, because it’s not the same type of investor.

First of all, when we talk about investors, there are really tons of different investments possible. There are plenty of possible investors. You have to choose just the one that best suits your model. And when you’ve just created a company, or when it’s been a few months or two or three years since you created the company, you don’t really know where you are, where you’re going too. And that’s where the biggest risk is, I think. And that’s where self-financing is still interesting. It depends on the reason why you create your company, if you create your company to make it work and be rich and be able to move on after fundraising it’s a very good way to be able to achieve this goal very quickly. Because the road as a self-funded company is long and painful, so it’s in the same way as saying I’m doing an open source business, but everyone knows by now it’s very hard to do it, it’s adding additional difficulties.

On the other hand, it’s how dependent you want to be or not in your supply chain, so you can redevelop everything yourself or you can rely on existing software or existing open source software, for example, but you take the risk of what it means and the controls you have to put in place. It’s the same for your money, you can also have in your technical and financial supply chain how I manage to pay people in my company, people from outside, and for the time being, it’s the same thing, you have to pay attention to this supply chain and know that you’re dependent on it. After that, it’s up to us to find out how much we want to be dependent or not. And so, we chose self-financing because that way, it makes us dependent on nobody. On the other hand, it is a brake on development. Today, I still have people who say to me, “Rudder, that’s great, you’ve been around for 15 years, but I’ve never heard of you, how come?” Self-financing is not easy.

And then, we’re in a niche that’s a bit special too. So, it’s not a hype project, we don’t deal with hype subjects, but just important subjects, in our opinion. And that’s the lot for a lot of companies, because we have customers like Banque de France, BMW, Free Pro, Samsung, BlackBerry, AFNIC, Soltec, it’s very varied.

So it’s not super hype either in itself and the fact of not doing funding means withdrawing a good part of all the press and all the free communication that comes with fundraising. And then also by the ability to bootstrap things much faster.

Alexandre Brianceau

So what it changes for us on a daily basis is that today I have the three founders me and we have an external board in addition. And we decide concretely what we do: the reality is that it’s not even with them that we really decide, it’s especially with our employees.

In fact, we have in a transparency approach, this co-construction, which is not too obvious now, by moving to 24, it’s other dynamics, but we always try to keep it active, which means that in fact we only know how to validate between us and even then there would be an external funder, it’s not even sure that he would understand this type of model and at any time he can say “do it”. more”. If a year ago it was the case, for example, of 2025, a new product, I’m leaving, concretely, I went on paternity leave for two months as a boss, I have my communication manager who left for six months on leave, so we were able to say to ourselves, we may be looking a little less for growth in turnover and we will rather try to build our commercial pipeline, so For the future, but with a financier who says me, in three years’ time I must have a payback [return on investment], it’s not possible, but there are some for whom it works very well, in any case we have freed ourselves from certain problems and we have created others, growth is slower.

Walid: Would it have been possible if you had looked for external investors to find them?

Alexandre: The founders took the step, in fact there was a talk that was made by François at the OSXP in 2019, something like that which talked about the approach they had taken. By the way, François would be very open to discuss, to do that again with you in another podcast.

There was that, at one time, especially because that’s what was advised. You know, when you’re part of startup studios or environments like that, bootstrapping, innovation, usually, you’re pushed to do that. And in the end, they wanted to stay on board. In fact, they didn’t find financiers who were going to be able to respect their approach as a responsible and sustainable company or who really understood where they were where they were going and so they made the choice not to raise funds and today that’s really the thing, we thank the heavens every day for not having done that.

Alexandre Brianceau

Walid: By the way, in the same conference you said another thing, I said to myself “ok, we really have to talk about it”: you said that at the time the founders had discussions with you because they were in the idea of saying “at some point we’re going to sell the company” and you weren’t in that idea. I understand that you weren’t in that idea and I find it very interesting because when you’re a founder at some point maybe you have the temptation to sell the company and get your return on investment

Alexandre: I would even go beyond the temptation that you may have the vital need, I think that all our listeners may have had cases of burnout for themselves or for others, well I don’t wish it on you but it happens. It’s an example like any other, it can be as fair as you can bootstrap a technology and have your first clients, but not to scale.

You may also feel uncomfortable with a managerial role. Today, we come back a lot from that, from the classic role of “I rise in the hierarchy and I do management in my life”. Fortunately, by the way. The problem is that when you’re the founder of the company, it’s not easy to have a role other than being a boss for a lot of people, both for yourself and for others, the externals, your employees that you’ve recruited. So yes, and in fact, it’s very cyclical for me, my life in the company too. I was an employee, I was appointed director and in fact, by appointing me director, I was told that the objective was to sell in the short medium term precisely so that the founders could move on and make themselves profitable. So, if you’re going to do that, you might as well take advantage of it to make it grow, but you had to choose the buyer well for the project to continue.

Me, on the other hand, it was I go up in management and I wanted to do more in the company, not to sell it but rather to put my stone in the building and that I thought that at the time there would be plenty of other people who would be in the same dynamic and that we could do great things. Nothing but it’s been 10 years that the company had actually existed at the time we were making between 4 and 600,000 euros in turnover on the product and it changed very little, we had 1, 2, 3 more customers per year.

And so that moment when I became CEO and said to myself, listen, I’m willing to look to see if we can sell, but that’s not going to be my primary objective anyway. So, I’m going to develop the box first and maybe that will give you the opportunity to resell at some point. And finally, everyone got back into the game. It worked very well and today, it’s no longer a subject on the table.

Walid: Yes, okay. This is no longer a subject on the table.

Alexandre: After that, we and I, personally, receive requests almost monthly from various players, both European and non-European, as well as US. Because we have a lot of customers in the USA and especially in the American defense. That’s the advantage of being in a niche, we have a good product, so that’s what people are interested in, but for the moment it’s not a big subject, but it’s sure that one day if there’s a big ticket on the table, maybe at some point, we’ll say to ourselves that we’ll go and do something else. Maybe with the same team and rebuild a new company, but why not. But in any case, for the moment, that’s not the subject and above all, it’s not the goal.

Walid: Do you have any subsidies?

Alexandre: Yes, we’ve had some. Now, we don’t have any more because we focused on our development, our growth and it went well and now, we will be able to do it again. It’s financial Mi’kmaq, but we didn’t have enough equity to ask for subsidies that were going well.

But so, yes, we’ve had them in the past, the first one was Pôle emploi at the time, which I think still exists, it’s France Travail, which allows you to benefit from your unemployment directly to create your business. So they benefited from that. We have the Île-de-France region which gave us money when we entered a program called PM’up. So each region has its own modules that allow different… They call it, I think, vehicles.

As a result, there are different ways of being able to put oneself on paths that allow you to obtain specific financing, whether loans or subsidies. So there, for once, I think it was the subsidy. The BPI had also given us a very advantageous loan so that was cool: I would recommend it to everyone a loan we get all the money out of your box it’s not a big deal and if it works well the loan was so advantageous that anyway it’s generally lower than inflation so you have a whole interest in borrowing and Systematic which is therefore undoubtedly one of the most important organizations for us today, attached to the Île-de-France region but which is a big driver. Today, it is an open source hub that is at the origin of the Open Source Experience today, well, always.

Walid: It’s the competitiveness cluster in the Île-de-France region, isn’t it?

Alexandre: That’s exactly it. As a result, they lead a number of companies around different subjects. So here, we are part of both, cybersecurity and open source. They have historically been open source and in recent years, cybersecurity. They organize events, they give advice, they give conferences. They can help in certain cases with the BPI or with, for example, France 2030, they were behind to help us with that. And then, because we are in France, I cannot fail to mention the young innovative companies, JEI, and then the CIR [research tax credit]. Once you’re over 7 years old, 8 years old, I don’t know, to the 8th year, so suddenly, in any case, it stops, that’s for sure. And it is giving us money to do R&D. And so that’s very good, because it drives innovation in France. and it’s true that today it’s still a significant part of the income we have in the company, well the R&D compensations we have and which work very well.

The community of Rudder

Walid: The next part we talked about it a little bit and the difficulties I wanted us to talk about the community of Rudder, I wanted to know what you knew about your community and how do you know it?

Alexandre: Not much. So it’s like I said, we’re an on-prem software, we do cyber. These are quite touchy subjects with in particular root or admin access on the machines. In short, overall, it’s an on-prem tool with no telemetry or anything active by default. The only information we really have on the use of the software is going to be the number of downloads, but which, everyone knows, is completely crazy because between the bots that download through a proxy but for 1000 nodes behind so in fact we don’t know anything, we only had one agent download.

Today, it is estimated that there are several hundred, certainly more than a thousand, or even a thousand users in the world, but we have no certainty. It’s just based on the download frequency and we discuss through a community chat, gitter which is linked to GitHub but is posted by Matrix, so I think or Element to discuss with the community. So now we have a few hundred people, I think we must be in the 400 to 500 people who discuss on this community chat, but we have to have an account that goes with all that.

Otherwise we have social networks. Redmine finally our issues part is on Redmine, the GitHub too in a way, a newsletter but in the end not that many exchanges with our community except on the chat and which comes especially when there are problems most of the time.

And otherwise they are our customers because our customers are also open source users in fact for a large part of their uses and so with them on the other hand we have frequent exchanges we build things especially with them. I’d like to change this user dimension a bit, but it’s spending a lot of money and time, development and reflection to embellish this community part and which today is not necessarily in the heart of the subjects for the company. We mainly try to deliver technological solutions to our customers first and which are quite representative, I think, of our users too.

Walid: You don’t have any notion of meeting every quarter or anything by video where people can join, that kind of thing?

Alexandre: We try to do webinars but we have very few people who come to register for the webinar so in fact at the moment I just haven’t found the right format that would please everyone. Are people really interested in it too? I don’t know to what extent this is the case, so it’s a real debate. If there are people who go through your podcast or if you want to make it a dedicated podcast one day with someone “how did I manage to really exchange with my community” with pleasure. But yes, it’s a subject that we’ll have to elucidate one day, but I think we’ll need to be about fifty in the company so that we finally have the means and the cubit width to do something interesting with it.

No, afterwards, otherwise, I meet a lot of people in the salons, things like that too. And I think. We had meetups and we did introductory days to Rudder as well at the time. In fact, that’s how I started to deploy my first Rudder in life. When I was working at Bull, I did an introductory day at Rudder. So, we stopped because in fact, especially post-Covid, it had already been some time since people had less and less time to test new products or things like that. They were pressurizing each other. So we stopped a little bit, but it would still be worth trying little things from time to time. Today, if I had to do something again, I would maybe start with a Reddit community, maybe a Reddit community, but not only focused on Rudder: maybe around security through Ops, but it takes time to animate this community, so there are people who listen to us and find it interesting and who don’t hesitate to contact me.

Walid: I’d like to take this opportunity to say hello to my friend Fabrice who I think wrote some of the first versions of Rudder’s doc

Alexandre: yes Fabrice Flore Thébault . yes outright who works on Red Hat I still believe now and yes yes outright the first iterations.

Walid: He was one of the first guests on the podcast to talk about these doc topics and everything and so in this community I was going to say who makes up the community, that is to say are there more kind of models with people who are in this type of organization or others? Or is it really hyper disparate?

Alexandre: Actually, Rudder is infra automation, so it’s very broad. There are also agglomerations, universities, university hospitals [university hospitals], industrialists, banks, outsourcing companies, schools. It’s really very varied. So that’s reflected in our community. They are only admin systems for the most part, at least who get involved. We have users who are on the security side, who are social security operatives, or very technical CISOs.

But most of them who are really heavily involved are more sysadmins or SRE or DevOps or whatever you want to call them depending on the fashion of the moment. And so, it’s more these profiles who come to exchange and who we have discussions and even more so with the customers so that we also plan the future of the product a little with them. It’s rare that we develop a feature without doing it jointly with a user, and so most of the time, it’s customers because they’re the ones who talk to us.

Yes, I said I would speak again. The fact that we have a business model that is not 100% open source in the sense of 100% free and the only paid part is because we have a SaaS offer makes me wonder, it’s a complete assumption, if people don’t remember their exchanges and discussions with us too much, afraid that we will come and annoy them commercially commercially speaking behind it. In any case, we have a really hard time communicating with our users as well, there’s also the case of prospects so people who ask us for information, we talk to them, maybe even do proofs of concept and then suddenly we have more news and in fact we find ourselves two or three years later learning that in fact they have remained open source users and they never dared to tell us because they were uncomfortable with wasting time and all that.

But we’re ok, well I mean, it’s part of the game and it’s a shame because in fact we’re cutting ourselves off from improving the product together, so it’s a shame, well, if there are ever users of Rudder Core among the listeners who don’t discuss with me or someone from my team, Please do it. We just want to continue to exchange, to expand the community in general. In fact, we have an ambassador program. We don’t ask for much. It’s especially that it’s full of quite advanced users who chat together.

Alexandre: There are people like Stéphane Robert, Frédéric Alix, Maxime Longuet, lots of diverse and varied people who are there to talk to each other, or even former employees like Julien Briault who works at Deezer today who moves for the Restos du Cœur for whom we give them a free version of Rudder for them complete who is also part of the ambassadors because they continue to contribute to their Rudder in some way and

Walid: To do conferences, I’ll put the link to the conference he usually gives in the transcript [How we transformed the restos du coeur into a cloud provider].

Alexandre: yes on the cloud of the heart yes

Walid: which is quite crazy as a thing

Alexandre: And force him because the number of conferences he gives is a machine, it’s good Julien yes and take care of yourself Julien if you listen to us.

Lessons learned from the Rudder experience

Walid: ok all this is super interesting but then the fact that Rudder has been around for a long time I can’t not do a part on the lessons you have learned from all these years well yes because in fact the thing is that precisely when you have a company that it’s been 3 years that you’ve been trying to find your market and everything but now you’re self-financed it’s been a long time since you are there you have seen a lot of things go by, you have certainly taken away a lot of things from it, so if you had some elements to give? What do you remember from all these years?

Alexandre: There are several, not on the same things. So already, making an open source business model is complicated.

What do I remember from those years? It’s that financing by making open source software, no matter which model you choose, it’s not easy. It is a job of passion above all, for founders, employees and contributors. I think everyone knows that, and we’re doing it because we believe in that model.

Alexandre Brianceau

We believe in the principle of the common good in particular, so we are very happy to see that Etalab, DINUM, all this is working well precisely because we need it. In fact, that’s my point of view but any software that the State uses should be open source in principle but it comes with counterparts and the business model and financing part is still not easy to do, you have to know it when you get into it I think everyone knows it but you can get out of it and it works just as well.

There are always companies that are created around it and so that’s it, that’s it, that’s the positive point and by the way, when you say is open source dead, I just think it’s making more noise because there are more and more big projects that find themselves closing their source, modifying their license, without saying that they close source but it comes down to the same thing with the BUSL for example or others.

Once again I’m going to do what you referred to in the talk I did on the subject, you have to know that there is a real issue of awareness for the months and years to come, it’s to raise awareness about licenses. Even the most passionate people are not necessarily open and it’s not just a subject of big free software to talk about licenses in fact it’s a real business subject in the sense of using a project without knowing its license and the impacts is taking big risks, you have to be very careful about that anyway and people don’t know it well enough.

Walid: That’s one of the objectives of the podcast too, it’s to talk about this and recently we had the opportunity to give several courses in engineering schools and the students were very interested in this bachelor’s side because in fact they didn’t understand anything, well they said “we didn’t know all that”, And in fact, your bachelor’s degree says a lot about what you do, about your project, where can it go, what you can do or not. So that’s also one of the expectations of the podcast, it’s to be able to talk about these subjects with people who master the situation and who are able to explain it.

Alexandre: yes definitely, and then you shouldn’t hesitate to talk about it around you, make shared lunches with other companies that may not have this knowledge.

The fact of simply popularizing things is not just a matter of technical choice, big geeks, I choose this distrib or that distrib, well if I make a parallel because there is always the war of “I’m on Fedora, oh no, I’m on Debian”. Yes ok but in fact licenses are much more than that and it’s actually much less technical than we think. It’s not easy to understand, but you can draw up some broad lines at least to get out of it and at least pay attention to it and especially alert your legal department to the risks associated with the company, because it’s often very badly managed by the legal departments of the companies.

Alexandre Brianceau

So business model is not easy… It’s cool to be on our own funds, we’ve already talked a lot before, so we’re not going to go back on it.

In the teachings, I would say that it was precisely when you told me it has been many years and so when you say it has been many years for me, it echoes the reality, which is that today we are in full swing of development.

For quite a few years it has developed but not at this level, so there have been triggers and among the triggers the first sorry for the anglicism is the product-market fit it’s the alignment with marketing with your market. It’s often techs who create open source software or companies so it’s obvious. However, we start with our engineering tech bias, let’s say in quotation marks, I feel it’s not the fact of having been an engineer or not but with this slightly technical vision of saying “I have a problem and here is the solution”. In fact, it doesn’t work if you don’t know how to communicate it properly and therefore if you don’t know how to market it correctly. And it’s not because you have nothing to sell if it’s a 100% open source and free product that you shouldn’t market it and think about this associated communication, otherwise you won’t have users and it’s still a bit of a shame.

Alexandre Brianceau

That being said, you can also just produce a code that you will make open source and maybe there will only be ten users in the world but that’s okay, at least you will have shared and that’s good.

But suddenly the product-market fit is the thing that I remember as something that may have failed to be as clear and speaking and not only to the people who are going to be the users, because in fact often if I additionally it requires resources, especially financial, but it’s also valid in time, you have to give the tools to the people who want to use you to be able to convince those around them.

By doing this, it’s a popularization work and it will also be useful to you in your growth because you will have to recruit people who are not in your field and above all that is looking for diversity, it’s super important, they will bring you a lot of things, these people and conversely also you will be able to transmit to them all your constraints and your knowledge today, I have a team, I really have a lot of people who are not at all tech or open source and yet who do business and marketing and commerce and who are doing very well today and we haven’t deviated from our original mantra and they remain very respectful of that, on the contrary, now they even see the interest in it, why we do these things, so there you go, we shouldn’t be afraid to Instead, you have to focus on people who are curious, who want to learn and who will be the driving force behind listening to you and learning from you. These are things that I will remember after all these years.

Walid: There’s a trick to it, but because it also comes from my own experience with GLPI at the time, it’s that the product is built over the long term and then at some point, indeed, it can take off, whereas for years, it can grow slowly and at some point, effectively, It can take off and that’s cool. But if your product has been built over a long period of time, it potentially has a good foundation. That doesn’t mean he doesn’t have technical debt, but he potentially has a good foundation. And so afterwards, the day it takes off, it’s cool, but already, there are a good part of the things that are built. The vision of the product is built. There are many things. And I find it really interesting to say that “it’s OK if you make a product and you try to make sure that in three years, it will have conquered the world”. But it’s also OK if your product, it’s been around for many years, it’s slow growing, it has happy users and everything. You have a diversity of things. I find myself a lot in this long time of the thing.

Alexandre: We often meet people who will say “When you have an open source product and I don’t understand, I don’t have many more users, it doesn’t take off.” Often, look at the product-market fit. Can it really speak to anyone? Do you manage to get people interested? Do you talk to them about it in the right way?

Let me take an example. From configuration management, we have shifted to a cybersecurity orientation. However, it’s the same product behind it, in the end. Of course, we have evolved the interfaces a little and all this to be in line with the situation, but basically, it’s the same principle for managing configurations continuously. It’s just the way we talked about it that made it take off. You have to sit down, take a step back and maybe even get people other than yourself to intervene on the subject for it to work.

Alexandre Brianceau

Without conquering the world either, it’s OK to say it’s a product that’s wobbly, but it can also be OK because in fact, if it takes off well, it will greatly increase the number of different user cases and it will help you see a little more clearly. At the beginning, if we have very few users, we can very quickly tend to develop the product to meet the challenges of one or two people and as a result we will only have these people and in fact we will go into one niche, we will put ourselves in even more niches, we will have a lot of different user feedback, even if the project is shaky and people do not remain users in it. Time because maybe it’s not stable enough but it will give you a lot of potential and interesting feedback to develop the product so in short everything is good to take in the end it’s never too late or too early to do something to believe in what you want to do.

Who to discuss topics around open source companies

Walid: All these systems, it’s not magic, it’s constantly evolving, etc. With whom, as the manager of a company, do you talk to about it? Are there other founders? There, you talked earlier about Centreon. With whom do you discuss these subjects a little?

Alexandre: So, as I said, I’ve been hanging around in the free software industry for a long time and I have my own network that has been created a little bit by little. Being a leader myself, it also opens doors for you to talk to other people. Podcasts are also super cool to meet other people in more depth for sure too. Through business too, I’ve met other people, but now it’s coming back a little bit on the executive side, but it also happens, so today there was that, I had, there’s a conference called the Open Source Founders Summit , which is done by Émilie Omier , which only brings together open source founders of open source software, which takes place in May, by the way, I think It’s next May 18-19 if I can ever afford to advertise.

Walid: She’s already on the podcast , we’ve already talked about it, she’s already done the promotion.

Alexandre: The promo is good, so it’s the people from Passbolt and Émilie Omier who created it. And precisely, since the people at Passbolt for the time being, especially Rémy [Bertot] and me and Émilie, we are rather aligned on the way we see the boxes and the open source project. And all of this overlapped, the objective was to have a place to exchange in complete transparency: so everything that is there stays there, we talk without complexes and without taboos about all the subjects related to the free world and to manage a company: so it’s also HR, it’s also licenses, it’s also business, it’s also marketing, it’s also dev, it’s really varied.

And then the Systematic competitiveness cluster which generates these opportunities like OSXP for example, so there are associated fairs, members of the CNLL of April , which also help and interesting meetings that allow us to discuss.

Otherwise, I don’t hesitate to look when there are confs that I like or people that seem to please me or that I don’t hesitate to contact them to try to take a little time to discuss, not always easy everyone is not necessarily very available, once again we are in times a little recession we will say in quotes so the noose is tightening and the availability Also but I like to discuss, discover new people, so it also helps to build a little, it’s great listening, we’ll be able to make a conclusion, we’ve been talking about it for 1 hour and 20 minutes already.

Alexandre: It’s true, time passes quickly.

Alexandre’s final word

Walid: I wanted to give you the floor for a final word, is there a message you would like to pass on to the listeners of the podcast?

Alexandre: Believe in yourself. Trust yourself. Take opinions and discuss around you. Be open source in your mindset too. And we have everything to gain by remaining transparent, and this is true for everything. I apply it with my friends, with my employees, with my customers, my prospects. Sincerity makes it only build good with benevolence and it won’t prevent you from being potentially disappointed or having bad experiences, but at least you will have done the best you could.

Walid: Very well listen, before we part, I would like to know if you have any recommendations for reading, listening or watching that you would like to share with our audience

Alexandre: I don’t have anything very special to share with you because suddenly I have less and less time to really discover simple things I must admit. In addition, with family life where I have two young children and so, if I have time outside the box, I used to read very interesting books.

Eventually, if I have a book, then it’s not at all about the open source part, so I come from tech, so not at all from business. So how do we learn things? I’ve listened to a lot of sales and marketing podcasts. I’ll leave you free to go and listen to this kind of thing. But in the books, there’s one that has a hyper-pragmatic approach to sales called John McMahon’s The Qualified Sales Leader. So, it’s in English, well, you have to read English and super practical for sales development. He takes a concrete case in cyber: I read it after a few years of experimenting with having done business and marketing. All the cases he presents, I’ve met them, so it’s very interesting to structure how to sell things much better. And then, if not, I have another one that’s in my bedside books actually somewhere… so this one is Scaling People – Tactics for Management and Company Building by Claire Hughes Johnson the same so it’s still in English but it’s super structuring for how we do it… I don’t like to go one again so I need to think about everything in advance and so this book helps a lot on that.

Otherwise it’s GitLab I think that has all their internal playbooks that are shared in open source and therefore are transparent and so if you ask about companies but it even goes as far as the remuneration model of the salespeople or the way they develop all this it’s in open access and that’s too cool too to learn things

Walid: great, I’ll put all this in the description of the podcast.

Alexandre, we’re coming to the end, that’s it, that’s it, it’s over.

Alexandre: Walid, that’s it, it’s over. We talk again whenever you want on other subjects.

Walid: That’s what I was going to say, I think there’s a lot to say.

Alexandre: With pleasure.

Walid: I think we’ll see each other again at another time. Thank you very much for taking your time to come and talk and share a little bit of Rudder’s experience and yours on this, I think it’s very interesting and really complementary with what we’ve already produced. It’s really cool.

For the listeners as usual, if you liked it share it around you and don’t hesitate to tell us if it inspired you for things or by the way don’t hesitate to contact Alexandre I don’t know Alexandre, how can we contact you if people have questions?

Alexandre: I’m on Bluesky , that’s my handle and then on LinkedIn or by email alexandre.rudder.io and that’s it, long live the Projets Libres podcast and then LinuxFr, you’re part of the free software ecosystem forever.

Walid: And there you go, you see, the long time, we always come back to it.

Alexandre: Oh yes, that’s clear. unbreakable.

Walid: The idea is to be here to last. See you soon Alexandre, thank you very much.

Alexandre: Ciao.

Episode production

• Remote check-in on January 12, 2026
• Editing: Walid Nouh
• Transcript: Walid Nouh

License

This podcast is released under the CC BY-SA 4.0 license or later