A small team of engineers and former practitioners who have seen the work from the inside, and decided the right way to apply intelligence is alongside the people doing it.
CEO and Chief AI Officer
Co-founder and chief executive of a European voice AI platform, after two decades running strategy and operations for international groups. Economics at ISEG Lisbon, MBA at UCL.
He leads Syntexia between London and Lisbon, from the strategy conversation with a leadership team through to the product that ends up running in their systems.
CTO and Lead AI Engineer
Co-founder of an AI health-tech product that reached 2,500 users, then Data and AI consultant to CAC 40 groups, where he took use cases from prototype through to deployment inside client teams. ESSEC and UC Berkeley.
He owns the technical side of Syntexia end to end. Every build starts with a refusal list: the things the model is not allowed to do. Deterministic engines handle the numbers, the model only writes around figures already calculated and locked, and nothing reaches a user without clearing an automated check. Each system is tested on both failure modes, giving a wrong answer and refusing a question it should have answered.
In regulated work, an almost-right answer costs more than no answer at all. These two hold on every build, without exception.
We are hiring for the work we cannot do fast enough. If you have done the work we automate, or you build systems other people depend on, write to us.