Ingram Technologies · Company in Belgium Tech

Overview

Ingram Technologies engineers production-ready AI systems including custom agents, LLM workflows, and models with GDPR and EU AI Act compliance built in from day one. The company offers security-first architecture with penetration testing, rapid deployment cycles, and privacy-by-design approaches for regulated industries including fintech, cybersecurity, and recruitment.

Key people

Jerome LeclancheChief Executive Officer

In the news

At the BSPHO hackathon #HackForGood with Google, the client asked for a chatbot. With our team at Ingram Technologies, we didn't built a chatbot. Instead, we built a fancy PDF viewer. Where every other team built a RAG-based chatbot, we saw the opportunity to dig into the source data, understand it, structure it, and offer a one-time AI-based pipeline coupled with a great app for doctors, rather than a chatbot that has privacy, cost and deployment implications. Our solution was the cheapest one to host and run by two orders ofLinkedIn · 2 Nov 2026
Europe doesn't need to build the frontier models to win" has been my pitch for the past two years and I was quite happy to see it reflected in Ursula von der Leyen's speech. This past week, TypeSafe AI showed a great example how to compete on form. Ingram Technologies as a lab also takes the approach of applying the research to focus on more efficient ways to train, deploy and use AI. But: Training our own models is not inherently a bad idea, we just cannot be trying to beat the frontier using the same tools andLinkedIn · 17 Sept 2026
QuotedDr. Corinne C-S@corinnecathI just listened to Ursula von der Leyen' EU State of the Union, and the #AI section left me confused. Let's work it out together. Her pitch is that Europe doesn't need to build the frontier models to win. The value lies in moving AI "from the screen to the real economy", onto the factory floor, the hospital ward, precision agriculture, the energy grid. And what powers it, she says, is European data, "a precious European treasure" feeding "tailor-made industrial AI models". It's the same bet Mario Draghi made in the FT last week: data is the one layer where Europe can still be sovereign. But take it apart, and three things don't hold together. 1/ "The European way" is asserted, never defined. "Tailor-made" hints at AI built without US frontier models, but she never says so, and nothing in current EU plans requires it. That silence matters. After this month's Navier-Stokes affair, where mathematicians alleged their unpublished work leaked to a frontier lab they were paying as customers, EU and US industry is newly wary of feeding proprietary IP into someone else's frontier model. For the "world-class industries sitting on the highest-quality data" she's courting, that's a real deterrent. So what does "tailor-made" mean? She doesn't say. 2/ The speech argues against itself. She plays down the frontier race, yet the €20bn InvestAI gigafactory scheme funds exactly that: frontier-model training. And it's mostly private money, only about 30% public. Who has pockets deep enough for the other 70%? Hint, hint: Big Tech. 3/ Furthermore, the "European way" is, in practice, not very European. The Cloud and AI Development Act (CADA) aims to at least triple EU data center capacity in five to seven years. But there's little native demand at that scale, Draghi himself admits it's so splintered it has to be artificially pooled to finance anything. So who fills the capacity? According SOMO much of it is financed by US investors and locked into contracts with the very US hyperscalers Europe says it wants to depend on less. A finding also echoed by by McKinsey & Company, it projects Big Tech will be 65% of European data center demand by 2028, in a reality where tree big cloud hyperscalers already control 75% of the market. From where i sit this means that every "real economy" deployment deepens our dependence on Big Tech, the same Big Tech von der Leyen treats as untrustworthy with our children in her under-16s social media ban. So its untrustworthy in the classroom, fine for the the operating room, the farm and the grid?
I've been coding for 24 years. 23 of them in open source. My very first open source contributions were to the Wine project. I wouldn't be where I am today without the open source community in general. But I recently realized that Ingram Technologies has not been the model open source citizen I want us to be. So I made it a priority to change this. A couple of weeks ago, we released a large amount of Financica as open source libraries. Today, we're releasing several new libraries as part of Ingram Cloud as open source, and thisLinkedIn · 23 Aug 2026
Your tools don't need replacing. They need to talk to each other. Composio's AI agents connect to dozens of external tools, but querying all of that as one coherent source? That's a different problem. We built the RAG layer that turns their connector ecosystem into a single, agent-ready data source; live, synced, and now live for their own users as a platform feature. "Ingram has built quality AI products and knows how to integrate super fast." — Karan Vaidya, CEO, Composio Full story → https://ingram.tech/case-studies/composioLinkedIn · 20 Jul 2026
Your company brain shouldn't go stale. Most companies document everything once, and then it quietly rots. New hires read outdated wikis, AI agents answer from last year's reality. Integrain keeps your company knowledge alive: always current, ready for your team and your AI agents. No migration, no ripping out tools. Curious what that looks like? We do free 20-min demos — DM Ingram Technologies or Jerome Leclanche and we'll set it up.LinkedIn · 11 Jul 2026

YBrusselsLocation

Something wrong or missing? Send an update. Fixed within 24 hours.

Overview

Key people

In the news

Related profiles