Inetum, leader européen des services numériques, a présenté lors du salon Big Data & AI à Paris, son nouveau framework d’IA : Coborg. Conçue pour accélérer l’adoption de l’IA et dépasser les projets pilotes, la solution a déjà démontré plus de 200 millions d’euros d’économies et 70 % d’hallucinations IA en moins. Avec cette innovation, Inetum entend offrir aux entreprises une transformation IA pragmatique, évolutive et centrée sur l’humain.
00:10You have been CEO of Inetem Solutions for one year, exactly today.
00:14Yes, pleasure.
00:15It has been one year, yes.
00:17I can see that you are focused on positioning Inetem
00:22as the preferred digital provider, especially in Gen.AI deployment.
00:27That's right.
00:28What is your point of view about deployment of AI here in Europe?
00:33Do you think that we are too cautious maybe
00:38or it's good to keep a certain distance, a little distance?
00:42Look, I think as far as Europe is concerned,
00:46I would agree with you that I think we are slightly behind the curve.
00:51If you look at the US, for example,
00:55just in 2025 about over 1% of the GDP growth has come from AI-related investments.
01:05Not AI as such, but just the investments that are going into AI.
01:08In Europe, I think we are not yet making that kind of infrastructure investment
01:14that is required to stand us in good stead for the future.
01:19And then in terms of just adoption, I think especially in mainland Europe,
01:24as opposed to say the UK or the US,
01:27the percentage of people that are using AI in enterprises,
01:33you know, consumer and enterprise AI are quite different.
01:36It's different.
01:37You can use chat, GPT, etc.
01:38Right.
01:39But when the enterprise uses AI, that is when it would be good for us as a continent.
01:43And I think that is lagging behind.
01:45We are probably at half the adoption as the rest of the Western world.
01:50Okay.
01:51Do you think that we are stuck in experiments here also in Europe?
01:56What is the biggest challenge in scaling up with AI?
02:01Actually, there are a number of things that people, enterprises think about.
02:05Right.
02:06And that is something that Anatom is trying to solve.
02:09And that is our mission.
02:11Yeah.
02:12The first thing is if you think about it,
02:14everyone talks about how you have to make a lot of investment in data
02:17because you can't use AI unless your data is well managed
02:23because it is all fragmented.
02:25It is unstructured.
02:26It is spread out.
02:27It's not consistent.
02:28And unless you solve that, you cannot use AI.
02:31So you need to first invest in making that.
02:33But we don't think that should be stopping the firms.
02:37The second thing is the security and the trust.
02:40The fact is, yes, AI hallucinates.
02:43That's the nature of the beast.
02:45That is how AI works.
02:46Because it's predictive, it is probabilistic.
02:49And the second thing is it's, you know, the security aspect.
02:52And I think in Europe, traditionally, that is our DNA.
02:57We like to be cautious.
02:58We like to make things very secure.
03:01We like to have a lot of regulation, which is good to some extent,
03:06but it should not inhibit that option.
03:09And I think that is happening to some extent today.
03:12But what I can hear is the challenge of finding the good use case.
03:18You know, maybe that's why we stay in experiments situation here.
03:24I don't know.
03:25But you are announcing today a new solution.
03:29That is the goal is to help companies to find real return with AI.
03:35How can you do that?
03:37In what way?
03:38Yeah.
03:39The new solution that we are announcing today is called COBORG.
03:42Yes.
03:43That stands for the cognitive brain of your organization.
03:45Okay.
03:46And what COBORG is, is a framework of various accelerators, IP, methodologies, and ways of adoption,
03:56you know, methodologies to make your people adopt AI, which sort of makes it, which is much more pragmatic.
04:02It is pragmatic in the sense that you don't have to make a lot of investment in data.
04:07We have a solution.
04:08So it's not a question of big money?
04:09It's not a question of big money.
04:10Really?
04:11Okay.
04:12So what we are trying to say is that, look, you can start small and then scale.
04:15And you can scale very fast.
04:17And that is what COBORG allows you to do.
04:19It allows you to manage your data.
04:22Like I said, people think I have to solve all this data and make a lot of investment.
04:25But we have a solution where we think you can use AI itself to solve some of the data issues that are used for AI.
04:34Do you have specific examples already?
04:37Yes.
04:38Yes.
04:39I mean, we are already speaking to a few companies and we are using it.
04:42We are customer zero for ourselves, for our own financial, you know, processes within Inetim.
04:50We've started using COBORG.
04:52And what we are seeing is that you don't, there are many things that you can do.
04:56If you can do small processes or small things that your people in the company can use.
05:02Yes.
05:03Then the adoption becomes much faster.
05:05You don't have to have a big bang approach.
05:07You can start with a small process, make that.
05:10And, you know, we have a methodology to find out where in your process should you use AI.
05:18There are some places where you can use.
05:20With a concrete return.
05:21With a concrete return with more impact.
05:22Really?
05:23Okay.
05:24You know, a methodology to say, are there some processes where you don't need to move to AI?
05:29Are there some processes where you can use AI plus a human?
05:33Or are there some other places where you can remain with the human?
05:37So that's, that's the way we are thinking about it.
05:40And we have a very mathematical and a very clear approach, which is based on logic that, that can be used and which can be done, which is automated.
05:50And it's adapted for any sectors?
05:52No, it is across all the sectors.
05:55It's all the sectors.
05:56Any sort of companies?
05:57Yeah.
05:58Well, I think obviously it would be very easy to use, for example, in financial services where if maybe you have a mortgage platform where, you know, you're talking about processes.
06:07Okay.
06:08It can be used in transportation companies.
06:11We are using it with some of our very large, very well known companies in the world in the transportation sector, where we are using AI to be saving them almost 200 million a year in their mission critical systems by using predictive AI.
Écris le tout premier commentaire