Leo Badigra

I split my time three ways. On Amazon's AGI team in Cambridge, I evaluate frontier AI models, testing how they respond, where they fail, and what that means for the businesses using them. Through my consultancy Tov AI Tech, I help companies and founders make sense of AI: what to build, what to buy, and where to start. And I personally build production AI systems, backend, retrieval, agents, and deployed apps. Most "AI experts" do one of these three. I do all of them.

Founders and small businesses figuring out their first AI move. Engineers and analysts who want to break into AI properly. Decision-makers who need someone to cut through the hype and tell them what's real. And anyone working in or with French-speaking markets who'd rather have the conversation in French.

Some version of "where do we even start with AI?" The answer is almost always the same: stop chasing the trend and start with a real, boring problem in your business. The companies winning with AI right now aren't the ones with the flashiest demos, they're the ones who picked one well-defined task and made it work end-to-end. I help you find that task.

ChatGPT will give you an answer. It won't tell you which 80% of that answer is wrong for your specific business. I work inside AI every day — evaluating it, building with it, deploying it — so I can tell you what actually works in production versus what only works in a demo. You'll leave with a clear next step, not a longer reading list.

Yes — and we should. Evaluating frontier AI models is literally my day job. I can walk you through what today's leading models genuinely get wrong, what risks are real versus hyped, and what your team should actually monitor when you deploy AI. This is one of the few areas where most consultants are guessing and I'm not.

Not at all. I work with founders, executives, and non-technical decision-makers all the time. I'll match my language to yours. If you want depth, we go deep. If you want plain English, that's where we stay. Either way, you'll understand what we discussed and what to do next.

Yes. I'm bilingual English–French and equally comfortable in either. Some topics — especially ones tied to your team, your culture, or your nuance — land better in your first language. Just let me know when you book.

Yes, within reason. Before AI became my focus, I worked across web development, data science, and cloud — and I still use all of it daily in my own AI projects. If your question touches AI from a data, infrastructure, or product angle, we can absolutely cover it. If it's purely an IT question with no AI component (e.g. setting up an office network), I'm probably not the right person and I'll tell you upfront.