Our services
Start from what you’re facing. Each service starts with something we hear from our clients. Find yours: what follows says what we do, and what you get. Everything starts small, on a bounded scope, and only scales on measured results.
Understand & train
Know where to start, and give your teams the right habits.
AI assessment and roadmap
You say — “Everyone talks about AI, I don’t know where to start.”
We do — Review of your processes, software, data and practices; map of possible uses (gain, effort, risks); choice of a first project and how to measure it.
You get — A prioritised roadmap, a first project ready to launch, and a frank opinion on what isn’t worth it.
Agentic AI training
You say — “We need hands-on training, not a lecture.”
We do — Short courses built on your own cases: stakes and priorities for executives; agents for admin work and documents for business teams; coding agents, tests and guardrails for developers.
You get — Participants who leave with concrete uses, and clear rules: allowed data, verification, confidentiality.
Automate
Hand repetitive work over to agents, under human control, so your teams can focus on their real work.
Custom admin agents and checks
You say — “We re-key orders from one system to another, and errors slip through.”
We do — Agents designed for your invoices, orders, drawings or work orders: they read incoming documents, pre-fill your tools and check consistency; a person validates, every action taken by the agent is logged.
You get — Less re-keying and fewer errors, business rules written down, and your choice of engine: cloud, European or fully on-premises.
Moving your development team to agentic AI
You say — “Everyone uses AI on their own.”
We do — Review of your current practice; setting up agents, conventions, guardrails and an automated test harness in your environment; working side by side on a real project.
You get — Measured speed with no regressions, a shared method, and a setup that is yours: your teams carry on without us.
Build
We develop our own platform, Visualia, with agentic AI: AI never works alone, our developers check and test what it proposes. We apply the same engineering to your software: code, tests and documentation are produced together.
Custom business applications
You say — “No software package fits the way we work.”
We do — Analysis of your processes, a feature register validated with your users, iterative development of a desktop, web or mobile application connected to your existing tools.
You get — Software that fits your processes rather than the other way round, tested and documented.
From critical spreadsheet to application
You say — “Our costing lives in a spreadsheet only one person understands.”
We do — Spreadsheet review with the person who built it: AI helps recover the buried logic; rules validated with users; rebuilt as a multi-user application connected to your ERP.
You get — A traceable, tested application, and documented business logic.
Shop-floor and field applications
You say — “The ERP doesn’t reach the shop floor: we write on paper, the office re-keys.”
We do — Applications on PC, tablet or terminal: forms, annotated photos, automatic reports, data exchange with the existing ERP, without replacing it.
You get — Data entered once, at the source, and real-time tracking of the shop floor or site.
Legacy software takeover and modernisation
You say — “Our software runs, but nobody dares touch it anymore.”
We do — Mapping the existing code with exploration agents; documenting the “why”; a non-regression test harness; step-by-step modernisation towards a maintainable base.
You get — Software that can evolve again, without regressions, and a backlog you can finally tackle.
Innovate
Bring the software and AI building block that turns research results into a tool.
Software & AI partner for research projects
You say — “Our consortium needs a digital partner that delivers.”
We do — Data acquisition and monitoring platforms, dashboards, interfaces to AI models, work-package coordination and deliverable writing.
You get — Research results usable in the field, and a partner experienced in collaborative projects: ongoing participation in European research projects (BFIRST, PVSites, BIPVBOOST, SEAMLESS-PV, RECLAIM-PV) and regional ones (Tremplin IA, Construction du Futur).
How we work
- Method — start small, measure, scale. No blind investment.
- Transfer — expertise and knowledge stay with you: no lock-in.
- Your teams — we automate tasks, not jobs: re-keying, sorting, checking. Your people keep the decisions and get time back for their real work.
- Confidentiality — on-premises or European options when your data must stay in-house.
- Funding — approved provider for the Walloon “chèques-entreprises” scheme; funding subject to eligibility.