2026.09.29 | TB | Software | Devoteam: Sovereign Cloud & GenAI Delivery Pivot

Analyst-led • 17 Sep 2026

Companies: CAP FR, NOW US

Carl Castanier (CC) Chief Information Officer at MidGard

Saad Maqsood (SM) — Third Bridge Sector Analyst

Agenda

• Devoteam's market positioning and competitive differentiation across European IT consulting
• Commercial execution and client adoption momentum across primary cloud hyperscaler partnerships
• Customer demand dynamics across enterprise and mid-market clients
• AI impacts on delivery model, offerings and potential revenue cannibalisation risks

Key Insights

• Devoteam's 2025 decline was driven by Broadcom's massive VMware product price increases, which forced enterprises to redirect budgets from T&M consulting towards licensing, disproportionately hurting Devoteam vs fixed-price competitors such as Capgemini. Switching from T&M to fixed-price contracts will be difficult for Devoteam, as it requires maintaining bench capacity and restructuring how talent is deployed and presented to clients
• Most large French enterprises lack the data infrastructure and budget foresight to execute meaningful AI transformations, meaning current AI projects remain small quick-wins rather than large-scale deployments
• Devoteam's recent AI expert additions are unlikely to increase daily rates, as customers are actively pushing for lower pricing, arguing AI makes consultants faster. Moreover, expert considers it to be "very difficult to save cost in Devoteam"
• Data sovereignty is considered an underappreciated upcoming growth driver, where Devoteam already holds strong credentials across sovereign cloud providers such as Scaleway and OVH alongside hyperscalers

Introduction

SM [00:00:01]: Welcome to Third Bridge's Interview today entitled Devoteam – Sovereign Cloud Realities & the GenAI Delivery Pivot. I'm Saad and today we have with us Carl, who is the former Consultant Expert and Director of Most Valuable Cloud and DevOps Architect and Technical Account Manager at Devoteam.

Just before we get started on the Interview today, Carl, could you please state I agree or I disagree to the following statement: You understand the definition of material non-public information and agree not to disclose any such information, or any other information which is confidential, during this Interview.

CC [00:00:46]: I agree.

SM [00:00:49]: Thank you very much. Could you start us off with an introduction to your background, please?

CC [00:00:55]: I'm Carl Castanier. I'm physical expert first and I switch on IT for working and I'm actually an expert of all distributed architecture. In Devoteam, I was an expert on Kubernetes. I have built a team with many certified Kubernetes people and the team grows very big, around 40 people. After, I work on business with commercial people in technical account manager for the big account, the big one like Thales, Safran because I know this market. That's all. I know all the IT except the development IT. I want to add that in Devoteam, I was a digital champion and this is a title for the better experts in Devoteam.

Questions

00:02:41] Devoteam's 2025 revenue declined due to various factors, such as soft demand and price pressure, across France, Belgium and Germany. At least in France, would you agree that the revenue decline was mainly driven by a weak market, or do you think the company might have faced internal issues that worsened the decline?

CC [00:03:30]: The issue was Broadcom for me. This is my point of view because they increased the price of their product, a very huge increase and it's not sustainable for enterprise and they have to switch funds from the time-and-material to the distance cost. This is only a transfer and this is because many companies don't anticipate the Broadcom buy from Dell and this is the main problem. For Devoteam, the impact is higher than, for example, Capgemini, I think, because Devoteam is more company in time-and-material instead of the fixed-price. Due to the time-and-material, it's more easy to cut people than cut projects or long-term projects or, for one, for example, you always need to run, so they cut on time-and-materials. From my point of view, Devoteam is not impacted like Capgemini or Atos or other ESN. The impact is lower because I can say after Devoteam, it's a learning company and in this learning company, there are many experts and expert is rare and precious, so the impact is lower when you are in high quality.

00:05:44] Would you say that it is tougher for companies holding more T&M [time-and-materials] contracts, such as Devoteam, than for companies with fixed-price contracts, such as Capgemini and Atos?

CC [00:06:03]: It's more difficult for time-and-materials, of course, because in fixed-price, you deliver in the time and you cannot cut the people. You have contracts with a long-term contract.

00:06:34] By the time you left, was Devoteam trying to pivot towards more fixed-price contracts? Is it difficult to switch from T&M to fixed-price contracts?

CC [00:06:56]: It's not easy to do. It's not easy because the model is not the same. In fixed-price, you have to have more people on the bench to be free for answer to customer demand. It's difficult to switch from time-and-material to fixed-price. It's difficult. This is the strategy of Devoteam. They want to go to the fixed-price because they want to have recurring funds with this business model. With fixed-price also, this is an effect. When you go to a customer, you have to present each worker to the customer and the customer accept or not the worker, but with fixed-price, you can get in back office some people who don't present well in face to the customer and they are very strong. You can hire very good experts, but not very good communication experts.

SM [00:08:37]: It seems like a good decision, though it might be difficult to execute.

CC [00:08:40]: Yes, after Devoteam grow a lot because when I come in Devoteam, we have 20% of growth. It was very huge. Actually, the CA, chiffre d'affaires, the turnover is more than EUR 1bn and for grow higher, you have to improve the fixed-price, for sure.

00:09:25] Devoteam's French revenue returned to growth at 4% YoY in Q2 2026. Do you think that quarterly growth is in line with the broader French market, or is the market growing faster or slower?

CC [00:10:03]: The growth of the market of IT is more or less the growth of the GDP, so it's 2%. They make 4%. Usually, we perform twice than the growth of the IT market. For me, it's in line of the objective, but they can obviously made better because they want to and the strategy is aligned to over-perform the market like they've done in the beginning of 2020 by the cloud because actually in cloud, Devoteam is a major company. If you want to run in cloud, actually, Devoteam have all awards in the four biggest cloud and also in sovereign cloud like Scaleway, OVH, etc. I think also that this company have four pillars, four entity inside. You have Impulse for strategic transformation for roadmap, for example. You have Innovative Tech, which is infrastructure and integration. You have Creative Tech for development and you have Cyber Trust for the cybersecurity. Their market is very spread, so some is higher one year, some is lower the other, so the result is the meaning of the all. Actually, cybersecurity climb very high and this is also correct to the AI because with the AI and sovereignty, the cybersecurity is very, very important and the market asked for these competencies.

00:12:51] Do you think Devoteam has the right service portfolio and expertise – such as cybersecurity and AI experts – to outperform the market, or does it need to improve its business model?

CC [00:13:49]: Yes, of course. Yes, because Devoteam, as I said before, is a learning company. A learning company is a company we skill regularly all employers and workers go to the office to the customer and show that they are up-to-date in skills. This is the main point in Devoteam that they don't stay on old technology. All employees, workers have to improve their skills during the time. You don't have any more network and system administrator in Devoteam. All is DevOps. All is automation. They improve to be at the right level in the technology, so all employees are very good experts. They can learn new technology and they like that. The frankness, passion and the other one is, I don't remember, this is the sentence. The passion on technology is it's main goal of Devoteam. In Devoteam, you can find experts will like their job and price to the customer is very good.

00:16:10] Devoteam has been going through a strategic transformation plan focused on AI-driven growth. Was this initiative already underway while you were at the company?

CC [00:16:46]: Yes, of course. Devoteam is always in transformation, always. The governance, Bentzmann brothers and Sébastien Chevrel, they lead the transformation of Devoteam, but the transformation is not now. It's always the transformation, always, always, always. They don't stay in the market. They always want to transform to improve the company. When I arrived in Devoteam, for example, their Amplify plan version one, version two and now, I transform. This is the ADN of this enterprise. For transformation, what they do, they take a good expert on old technology, they hire him and after they can reskill him on new technology because he's a good expert on one technology, can be a good expert on another technology. This is a way to transform by the time. Also, this governance give a way. When I was in the team, I think it's an authorisation way because they said the market is there and everybody going in this direction.

For example, in cloud, it's worked very well. For AI transformation, they make many things very good. They skill all people on AI. What is an AI? What is the ethics with the AI, how you have to use it, how you can have the better answer of an AI? They do any meet-up for any RETEX, for example. They try to have some partners which include AI in this technology and provide services on this technology. For example, you have GitLab with AI, you have Palo Alto with AI, you have ServiceNow with AI and each domain where is Devoteam, we have to find something to do with the AI and more or less with the time, some experience growing and after the Devoteam becomes a leader of the market like they do in the cloud, so transformation is the ADN of Devoteam. This is probably the only one company where I work that I saw that.

00:20:46] Would you say Devoteam was leading the market in AI solutions by the time you left, or were other players such as Capgemini, Atos, Onepoint or Talan better positioned?

CC [00:21:16]: Actually, there are no leaders in ESN AI. Devoteam is not a leader. Accenture is not a leader. They have some competencies inside, but this technology is new for the emerging enterprise. It's not new at all. It existed since many years, but the introduction in enterprise is very new. The point of Devoteam is that before they work mainly on roadmaps, on enterprise architecture, so they know what is data governance, for example. They know how to use data, how to securise data and this is the point that it can be an accelerator for the future market. Actually, on AI, all customers try to look for what they can do with AI with the legacy because the legacy is it's a pain point for AI. No API, no possibilities to get data, it's very difficult to have a very big AI project. AI project actually is a very small, quick win for customer, so it's a small project and sometimes the visibility of this project is very high and the customers don't want to fail. Devoteam work with high quality, so I think in markets, they have a good pace because with experts, you have more chance that you don't fail.

00:23:41] As part of Devoteam's latest transformation plan, restructuring efforts and associated costs have been reported. Were these restructuring costs driven primarily by hiring AI experts or by reorganisations, business model restructuring and other expensive initiatives?

CC [00:24:28]: More or less the mass hiring is always a trend of Devoteam within, always hire and hire. You are the boss. Hiring is business as usual. It's not a problem. From the customer point of view, because I was also a customer of Devoteam, the acceleration what I see is that, I talked before of four pillars, cybersecurity, Innovative Tech, Impulse, etc, and actually, if you deal with Innovative Tech, for example, you have a big chance to have contact with Impulse for strategy, Innovative Tech for development, cybersecurity with Cyber Trust, so they want to maximise their customer. I want to say that when one entity work with a customer, they want the other entity can start a new market and start a new customer in this entity. This is more or less the Amplify plan of the beginning of 2020. This is this transformation. It's not very easy because all entity is segmented and he will make some time to do that and to finish this transformation, but they do, they do, they do. I can see that.

00:26:44] Do you think the main focus of Devoteam's transformation plan is to increase revenues or to cut costs and improve efficiency?

CC [00:27:24]: Clearly, Devoteam want to increase this revenue. Clearly, they want the saving cost in Devoteam. It's very difficult to save cost in Devoteam, I think. You can save cost on the restaurants, for example, non-business usage, but in Devoteam, I think it is not a subject. I don't see that. They want to improve the share of the market always and go where they are the more grower of technology. They go to AI not because they love AI. I think I talk about the governance. They go to the AI because the market is on AI. They go to sovereignty because the market is on sovereignty. When they see that the market is there and Devoteam is not, they want to go. It is only that. They want to increase the revenue. From my point of view, it's not a good thing to want to reduce costs when you have a revolution around you. You have to try to have a good cart and go into this market. I'm not sure I'm clear, but Devoteam want to increase the revenue and grow to EUR 2bn. This is what they want, I'm sure. If you want to reduce internal costs, you can grow. It's not possible to reduce cost and invest. Investment is at the end of the Devoteam and Bentzmann brothers. For me, I saw them like that.

00:30:15] You said there isn't a clear AI leader in the market yet since all companies are still figuring out how enterprises use AI, but do you see any player who's currently best positioned to deliver AI projects – whether Devoteam, Accenture, Capgemini or others? Specifically in France, which players are outperforming their peers so far?

CC [00:30:59]: For me, it depends of people, mainly more it depend on the people than the company. On a project, if you have a good people, the right one can be a successful project, but it's not due to the company actually because few people have experience of AI. In Devoteam, they have some experience because what I said on data, we work on data. We work also on from emerging some bots, for example, in companies. We perform AI before, but this is not the AI like we have always now. The AI actually, if you see what I can do with ChatGPT at work, it's incredible. I can say that I'm 3x faster than before. It's incredible how the AI accelerate my work and I don't work less. I work more. It's not the same because before, I do something with lower quality because I have to do it fast. With AI, I can do the same job with high quality and with the time, so I have time to do other things and all quality is improved. It's not the same than before. It's not the question, I see. I go on the other topics, but it's just for illustrate my proposal.

00:33:10] Devoteam is always hiring talent, as you mentioned, and the company recently hired AI experts, which will hopefully boost its pricing. Were these AI experts already integrated into the business by the time you left or do you think this occurred afterwards?

CC [00:33:46]: This is the way of Devoteam to transform the company. They hire some experts, a few on because they are very expensive and after these experts have to reskill all the company to be ready to go to the market because if you're hiring only high experts, your price explodes. It's not possible to have this kind of business. The idea is to have AI experts, which are able to help people who have some leads, for example, or some issue. I think it's in decision champion community because the decision champion community is more or less 200 people, I think. I don't sure the number, but they are the good experts. All is very, very, very strong in their position and you can ask them an issue and they have time to answer and help you for your business. This community, I think, is very important for the future and for the transformation plan.

00:35:36] Does assigning an AI expert to a project command a higher daily rate? How much could Devoteam increase its price by adding AI experts?

CC [00:36:04]: For me, you cannot increase the price because actually, the market is not falling for the price. Customers want to decrease the price because they said that AI make a job for them. This is a decrease of price, so AI expert is at the same price because they are skilled on AI, so they don't reduce the price. Do you understand?

SM [00:36:44]: Customers want lower prices because they think AI allows consultants to do the work faster.

CC [00:36:54]: Yes. The idea is if my expert is lower price, I skill him to AI, so now, it's not an expert. It's an expert with skills to use AI. The point is when you use AI, you have to have ethic and you have to use it in the right way. For example, I saw in customer that someone asked to an AI to perform some job and the result is a monster and this is a very big huge problem actually. It leads to issue that nobody understands what the AI build. After when it's in production, you have to rework and understand AI is not easy. Good usage of AI is sold more or less by Devoteam. This is a way to keep price at the right level. I think it's difficult. The market is difficult actually. Journey is difficult.

SM [00:38:31]: As you mentioned, AI is still very new, which makes it difficult to set pricing for these projects.

CC [00:38:42]: It's difficult to price very high even if you have AI experts.

SM [00:38:48]: Customers won't want to pay more for AI projects. They want to keep the same price or lower.

CC [00:38:52]: Yes.

00:39:00] Over the next 1-2 years, while AI is still new to enterprise customers, do you think Devoteam's margins might go down due to pricing pressure and the high cost of its own AI tools, combined with customer demands for lower prices?

CC [00:39:38]: I'm not afraid of that because revolution is the way IT always improve. This is my opinion. I think in the future, you will have more work with AI. AI will destroy some jobs for sure, but it will create the new one and we can do things that before we cannot do. One in one, you have a new business, you have to go to the new business and don't want to embrace this new technology. For example, you have AI agentic for hyperscaler, which can replace many technical people, not expert people but technical people for monitoring, for example. You can have very aggressive decrease of cost, but to put in place the solution, you need an expert and when this solution don't work because always there are, for example, you can have hallucination, you can have a problem with network that the AI don't know, so you always need people to put this in place.

I think, more or less, the market will grow in the future and AI is not a problem for me. Just AI will lead, the word in English, to obligate worker, to work with a very, very high quality because actually, when someone delivers something to me, I put it in AI and I saw in 10 minutes what is the problem, so the quality delivery will be higher and the job will increase because we will focus more on the business than the IT infrastructure and all things that are not focused on the business of the customer. For example, the go-to-market will grow with AI and this is the better benefits, I think.

00:43:27] Regarding AI's impact on contracts, are T&M contracts affected worse than fixed-price contracts, or is the impact of AI the same regardless of contract type?

CC [00:44:05]: I know an answer, but I don't have the yes or no. I put my position on the table. For time-and-materials, you are behind the customer. When you use AI, the customer know that. When you are in fixed-price, you are not face to the customer. The deliver is face to the customer. In fixed-price, customer probably make more pressure on the price because they can so that it's possible to perform this work with the right AI with the right parameters. I think for time-and-material, the impact of AI is an added value and for fixed-price, probably a higher pressure on the price. This is my reflection, but I don't know we can debate on that.

00:45:43] Do you think hyperscalers such as Microsoft, Google and Amazon could be a threat to traditional IT services companies such as Devoteam by taking business away due to GenAI, agentic AI, etc, or will they remain key partners going forwards?

CC [00:46:20]: Devoteam's point of view and actually, Devoteam is on build. The Devoteam business is mainly project and the impact is lower on project because agentic is more something when it's in place, you don't need people to watch screen and monitor or perform manual action, so the impact is more for company which perform the run. On Devoteam, of course, you have a little bit run and there are some impact, but on build, the activity will grow because you have to put in place the solution of hyperscaler and Devoteam is very strong in hyperscaler, though. I think they can improve the business on the same position that they have already with the customer. I think it's not a problem for Devoteam at all. Where I work in Devoteam, I was in sovereignty cloud and innovation, so I know less the cloud entity, energy cloud and cloud. I know less, but I think because I know some people inside that they have all skills to improve in agentic and take shares of the market in this space and probably my structure will decrease, but if they grow and decrease, that also sounds good.

00:48:57] People think the software side faces many risks with AI, but when it comes to IT services, do you see any risks for companies such as Devoteam due to AI or any big changes they have to make to their business models?

CC [00:49:25]: It's a problem for legacy company. I don't know what's happened in Atos, for example, but for me, it's a legacy company. They don't transform the company and reskill people like Devoteam, so the transformation of Devoteam is only a way to don't be locked in old generation and always to go to the new one. I think Devoteam is not an ESN that will be impacted by AI and probably it's a way for them to improve the share on the market because legacy company will have very, very big problems, I think. They will have problems also. When I see Capgemini in India, probably most of Indian employer will be fired because of the AI because they use this the low cost in India to make manual action and this manual action will be cut by the AI, so the impact will be higher for a major company. This is my point of view again.

00:51:19] What is customer demand for AI like in France? Are enterprise clients still in the exploratory proof-of-concept stage, or are they committing budget to large-scale AI projects?

CC [00:51:43]: AI is a communication project first. You have two things. All innovation budget go to the AI, for sure. Now, the innovation budget is AI and nothing else. You put it by side. Also, you have traditional IT which is more or less AI. For example, I have ServiceNow. I deploy an AI service and I said I do AI, but it's the same product. The traditional IT budget switched to the AI project, but no company have overseen the AI transformation, so the PMT is not prepared for the AI and the budget of the AI because if you want to do AI, you need to put money on the table and not a few money, very, very huge amount of money. For example, if you want to perform deep learning, the Nvidia cards is very expensive, and you don't need one but many. The budget is not, the word in English, foreseen two years before. Probably in the future or next year, the PMT of all major enterprise will have a big amount of budget for the AI, but actually, I don't see a big, big budget for the AI. They want to do something, but they don't know what they can do with their traditional IT.

Now, they have to transform their IT. When I was in the Devoteam, I said to my colleague, "Enterprise must be that ready to be AI-ready." It's very important to understand that data is the gasoline. It's the petrol of the AI. If you don't feed the AI with the good data, your AI don't perform like you want and you can lost many people and have bad experience and bad experience in enterprise. It's a delay for the future because you have a defiance of the technology. If you do that, it's a good way to fail the transformation of your enterprise. First, data-ready and AI-ready and I think bigger major is not ready and Devoteam work mainly with major enterprise in France, the 40.

SM [00:55:58]: They represent the 40 largest French companies.

CC [00:56:03]: SBF also but mainly CAC 40 because it's better to work with a big company. You have more money.

00:56:24] Devoteam has strong relationships with Google, Microsoft, Amazon and ServiceNow, which account for about 33% of group revenue. Are these strong relationships an advantage for the company's AI readiness, given how AI-first these partner companies are?

CC [00:57:05]: This is what I said before. Yes, it's an advantage because they are in place in customer. It's better to be in place to look for than look for a new customer. You are seat in place and you just have to discuss with your customer to improve what you already do. The relationship between the customer and Devoteam is already done and I think customer is very happy of the Devoteam work, so it's more easy to solve something and AI with cloud provider. What you read, it's true. Devoteam is very strong in ServiceNow and cloud providers, very, very strong. For me, this is actually, the best company in cloud is Devoteam for me, small one but the best.

00:58:44] Under its Boost Amplify strategic plan, Devoteam targets EUR 2bn in revenue, with nearly 50% tied to AI projects by 2028. Do you view this 50% target as ambitious, realistic or potentially conservative?

CC [00:59:36]: This point it's complex because in Devoteam, the more important is not EUR 2bn or EUR 3bn. The objective is the way and as I said, it's an authorisation way. Everybody want that the strategy will be realised, so everybody will do the best to reach this objective, which is important. It's not to lost share by making projects which fail. Devoteam have to keep the image, the logo of good quality. Devoteam want to take his image like employer, which is the best on the market and deliver at the best. We don't want to bring some fails or something. I think customers are very happy to work with Devoteam. When I was in Safran, for example, Devoteam was my provider. When I want someone very specialist and very good, I go to Devoteam and this image is very important for Devoteam more than the success of price. I know that you want to know if the price grow, but the image for me is very important because it will lead Devoteam on the long-term project. It's not a short-term project that in '28, we stop all and we stay on AI.

I think in one year or two years, Amplify will be obsolete and the governance will make a new strategic plan. They will switch to the market where the market is. The way is not strong. They just want everybody go the same direction and if the direction don't work, they can switch. In this plan, some things will perform, something will not perform and they have to take a lesson of that and after make another plan and so on. It's very difficult to say that the plan will be finished or will apply with 70%, very difficult for me to say that. The whole objective of this plan is to put all people in the same way and try to do the best to wish a success and for me, this is a good way to manage an enterprise. It's very good in Devoteam, Stanislas talk a lot and it's very important for each worker to understand what your manager said and it's aligned on what you think on the market. They are very strong on that. In my life, it is the only one company that I was aligned with my president.

SM [01:04:30]: It's a good thing.

CC [01:04:35]: I'm not clear. I have difficulties to say what I see.

01:04:46] Is there anything else you want to add about Devoteam that we didn't discuss?

CC [01:05:14]: One thing that we don't talk is sovereignty and I think Devoteam have skills to improve in sovereignty. I think this is the next issue in the world is where are your data and Devoteam is at a good place and I have good skills to help company for that.

SM [01:05:50]: Thank you so much for your time today, Carl. I really appreciate it and I found that incredibly helpful and it was a pleasure speaking with you. I hope you enjoyed it, too.

CC [01:06:00]: Yes. Thank you. Bye-bye. Ciao.

SM [01:06:28]: Ciao.

Transcription ends at 01:06:28 of the recorded material