Overview
Founded in 2013 by Nicolas Deruytter and Michael Lemmer in Ghent, ML6 operates as a strategic AI engineering partner for enterprises across manufacturing, financial services, retail, healthcare, and government. The firm takes companies from AI experimentation to production, covering data engineering, model development, deployment, and MLOps, with strong specialization in Google Cloud infrastructure.
In 2024, ML6 became an OpenAI Services Partner, broadening its capabilities into enterprise generative AI engineering beyond its cloud-native foundations. It has been recognized in the Deloitte Technology Fast 50 and the Financial Times FT1000 list of Europe's fastest-growing companies, reflecting consistent commercial momentum.
ML6 employs around 136 people across offices in Ghent, Amsterdam, Berlin, and Munich. The firm has become one of Belgium's strongest AI talent pipelines, with alumni having founded a notable cluster of Belgian AI startups including Paperbox, Uman.ai, Adomate, Leadriver, Powernaut, Scope AI, and others.
Media & appearances
- SuperNovaYouTubeSuperHero Conversations EP6 | Pieterjan Bouten - Entourage & Stan Callewaert - SquireAI consultancy where Callewaert worked, moved from technical to sales roles, and whose alumni network (including his DPO Rob) inspired and supports his entrepreneurship.
Key people
1 alumniwhere they went →In the news
- A two day old forecast just beat a three hour old one. Google DeepMind released WeatherNext 3 last week, its most advanced weather AI model yet, refreshing forecasts every hour instead of four times a day. We wanted to know what that actually means for renewable energy, so we ran it head to head against ECMWF's IFS, the industry standard for medium-range weather forecasting. We tested the three variables that matter most for a renewable portfolio: → Wind at 100 meters → Solar irradiance → Temperature WeatherNext 3 came out ahead on
- Next week we're heading to HumanX Amsterdam with our partner Google, and we'll have a spot at their booth. While most AI conferences try to predict what's next, HumanX focuses on what already works. That's exactly the conversation we want to be part of: how AI ambitions turn into systems running in production, drawing on the 150+ AI projects ML6 has delivered so far. We're also taking the stage at Google’s booth twice: → September 23 & 24: AI-Native Cybersecurity, with our Managing Director Wouter Desard. Sebastiaan Tuts, Kathy
- Next week we're heading to HumanX Amsterdam with our partner Google, and we'll have a spot at their booth. While most AI conferences try to predict what's next, HumanX focuses on what already works. That's exactly the conversation we want to be part of: how AI ambitions turn into systems running in production, drawing on the 150+ AI projects ML6 has delivered so far. We're also taking the stage at Google’s booth twice: ◾ September 23: AI-Native Cybersecurity, with our Managing Director Wouter Desard. ◾ September 24: AI-Native
- September is for new beginnings. For some, that means a new school year. For thirteen people, it means a new chapter at ML6. We're growing this month across marketing, HR and AI engineering. A full class, as we like to call it. Join us in giving a warm welcome to Hanna Vande Vyvere, Alexandra Dolbeau, Efe Atikkan, Niels Tiben, Wannes Dewit, Victor Geneste, Ellen Flamée, Foudil Brétel, Revaz Lobzhanidze, Aino Eskelinen, Markus Sass, Hashaam Hassan and Alexander Amin. Welcome to the team. We're glad you're here. 👏 Meet them below 👇
- AI-first engineering workflows, live in Brussels on 28 October. Co-hosted by TechWolf and ML6. The theme is the real working practice, not the demos: spec, build, test, ship. Three talks from teams doing this in production, a walking dinner to open the night, drinks on both sides of the programme. We've got a full talk on the lineup: ▪️ Niels Lowie (TechWolf) on what Specto builds overnight, and what still breaks. ▪️ Christophe Geudens (Prismos) on two years of agentic coding, including what it did to code review. ▪️ Georges Lorré
- Every year we hear about a good AI idea that never gets built. Not because the tech is missing, but because there's no time, budget or team to take it from idea to prototype. That's the gap ML6 for Good is meant to close. We look for meaningful challenges where AI can make a real difference, for people, nature and society, from healthcare and social welfare to environmental protection. It works in three ways. ◾Internships for Good pairs an ML6 intern with senior engineers on a project with real impact. ◾Christmas Projects is a
- From pilot to scale: how an AI assistant got faster and more accurate. Response times cut from 8 seconds to under 5. That is one of the changes behind 𝗛𝗲𝘆𝗜𝗺𝗺𝗼, the AI assistant helping people navigate ImmoScout24 every month. ImmoScout24 is where around 19 million people a month work through moments like that: searching, comparing, valuing, financing, buying, or renting. HeyImmo was built to guide people through it. The hard part was scaling it up without losing speed or quality along the way. Our team at ML6 worked with
- Why the cheapest AI model can end up costing you more. Most teams pick an AI model the way you'd pick anything else: find the cheapest option, use it, move on. We ran into a case where that logic falls apart, and it's worth understanding why. We tested OpenAI's new model, GPT-6 Astra, against its predecessor on a real task: migrating twelve complex data pipelines in one go, over 300 files. Astra costs 2.5 times more per token to run. Sounds like the expensive pick. But when we ran the older, cheaper model on its lowest setting,
Open roles
- Forward Deployed EngineerGhent · Aug 28, 2026
Alumni 9 went on to found or lead
Andreas De NeveAI Engineer Intern→Co-Founder & CEO at TechWolf
Florentijn DegrooteSquad Lead & Energy Domain Owner→Co-Founder & CEO at Powernaut
Frederic StallaertKey Account Manager→CEO & Co-Founder at Paperbox
Jules Van ReemptsMachine Learning Engineer→Co-Founder at Wintro
Kobe IndemansEngineering Intern→Co-Founder at Recommand
Lucas DesardSenior Machine Learning Engineer→Co-founder & CTO at Adomate
Milan PelgrimsTalent Partner in AI consulting (Via In Five Years)→Founder & Talent Partner at In Five Years- Simon LoggheManaging Director→Co-Founder & CEO at Adomate
Stan CallewaertCommercial Lead Solutions→Co-founder & CEO at Squire
Charles BoutensMachine Learning Engineer→Advisor at Wintro
Brecht CogheSquad Lead
Related profiles
Something wrong or missing? Send an update. Fixed within 24 hours.