What we are building
The R&D investment goes into:
- A chat interface for administrators to manage and steer AI activity
- Automatic workflow reconstruction, so agents can map and rebuild business processes themselves
- A secured master platform on which multiple AI agents can work together safely within business environments
- Closer collaboration with the Flemish company ML2Grow on specialised machine learning challenges
It is the step from classic generative AI, which waits for a prompt, to agentic AI that understands a task, makes a plan and carries it out. Think of agents that independently build landing pages, translate site content, launch marketing campaigns or monitor KPIs around the clock.
Or as Dominique puts it:
An AI agent that does not know your specific business context is nothing more than a nice demo. And an AI agent that works without safety rules or approvals is a real business risk. We provide the necessary control, so companies can deploy AI with peace of mind.
A quality label from VLAIO
VLAIO's independent experts assessed the application on substantive innovation, technical feasibility and market potential before approving it. According to them, the project “goes beyond applying existing technology”, and real research and development is needed to make autonomous AI agents work together safely within business environments.
The assessors also pointed to the combination of open source, autonomous agents and strict data sovereignty as a real competitive advantage in a market where there is still little on offer. They also mentioned Dropsolid AI's strong Flemish roots, through partners such as ML2Grow and Upgrade Estate.
Real results already
This is not theory. Dropsolid AI's approach already delivers measurable impact today:
- EPSO, the EU's personnel selection office, saw the number of support tickets drop by 90%
- Pidpa, the Flemish drinking water company, uses the sovereign answer engine ‘Pip’ to help more than a million citizens find the right information faster