Deployed engineering for SMEs
I help SMEs solve their most costly operational problems with AI, starting from their real-world constraints and deploying custom systems that work in production.
What I do
Deployed engineering is how AI gets into production inside an SME.
Instead of starting from a generic product, I build custom AI systems directly inside the complexity of your company: where Excel acts as a database, where tools do not talk to each other, where field teams still write things down on paper, where no one internally knows how to code, and where every euro spent has to justify itself. These are not blockers. They are the reality the system has to work with. The goal is to solve the problems that cost the most, under real conditions, with measurable impact. To move from "we know AI exists" to a tool that runs inside your business and improves profitability.
How I work
I step in when the problem is clear but no existing solution fits your reality.
The best AI systems are built from the problem, not from the technology, with one absolute priority: speed and real impact. I work with the owner, managers, and field teams to understand how problems show up day to day, then I ship quickly and improve continuously. The goal is simple: AI systems that work in the real world. Not in a demo.
Case studies
RGMC France
RGMC France managed recipes, batches, stock, production, and documents across several tools. I built a business application that structures recipes, tracks lots, manages composed ingredients, organizes R&D folders, and produces production and R&D reports by week or month.
+200 recipes synchronized, traceable, and reusable inside a deployed production system.
ROBOTIQUE ELEVAGE SARL
ROBOTIQUE ELEVAGE SARL needed a tool to help its experts analyze dairy cow rations without handing the calculations to AI. I built an assistant where Claude proposes hypotheses, the application calculates nutrition balances deterministically, and the expert keeps the final decision.
A controlled AI workflow: AI proposes, software calculates, the human validates.
RGMC France
The RGMC France team worked across several business tools that did not always share the same information. I built an operational agent connected to their existing system: client creation, sales follow-up, payments, reminders, automatic delivery-note and label printing, and client knowledge rebuilt from internal data.
19 operational workflows structured around an agent with human validation before critical actions.
Let's talk about your business.
30 minutes to understand your situation and tell you what AI can concretely change.
Book a call