Florian Bruniaux / About
Fractional Chief AI Officer
I like understanding how things work.
I ask a lot of questions. When a tool makes work harder, I look at how the team uses it, dig into the code when needed and test a solution with them.
01 / Experience
Code, teams
and decisions.
- Engineering Manager
- VP Engineering
- CTO
- CTPO
- Team size
- Approximately 30–40 people
- Management
- Multiple squads · Managing managers
- Organisation
- Distributed across several countries
At Cheerz, I went from Engineering Manager to VP of Engineering. I worked with the CTO to set the technical strategy for infrastructure, backend and web.
At Native Spaces, I went back to writing code daily as CTPO. I led product design and helped rebuild the web and mobile platform with a fully remote team.
At Méthode Aristote, I took a WhatsApp prototype to a first tutoring platform. I set up the checks that allowed product colleagues to contribute code.
Explore my professional background →02 / Build in the open
Building with AI.
Contributing to open source.
I became active in open source in early 2026. AI lets me explore and publish tools alongside my client work.
I enjoy taking an idea far enough for someone else to use it, find its limits and tell me what I missed.
I define expected behaviour, test changes and review the implementation. I remain responsible for the code I accept and deliver.
Claude Code Ultimate Guide
The guide I write and maintain alongside my engineering work.
ctxharness
I build this tool to detect drift between agent instructions and the codebase.
RTK
I joined Patrick, RTK’s creator, early on and contributed heavily to the code. I now focus on spreading the word and follow the project closely.
Big up to Patrick, Thomas and Adrien. Probably relieved that their most talkative supporter is busy elsewhere.
Explore RTK Pro ↗03 / Share the practice
Helping people develop
their judgement.
I have recruited, trained and supported many dozens of developers, from juniors to senior engineers. I adapt the depth and pace to the person in front of me.
Whether I am talking to a developer or the person funding the work, I want them to understand the reasoning well enough to question it.
When working with an agent, I often come back to diagrams: I ask how the blocks communicate, then have it explain the one I want to challenge. I try to recreate the discussion we would have at a whiteboard. I describe this way of working in my recap of the Insitoo live.
I offer my own training and speak about AI engineering and adoption. Find my talks and interviews or explore training and mentoring.
Earlier community work: tech meetup recordings.
Curious, and usually working on something.
People who know me recognise the tall bald guy with a smile. I ask questions, listen and say what I think, including when an idea needs challenging.
I like understanding why a tool behaves unexpectedly, even when my own tests have misled me. While building my memory observer, I wrote tests that gave the model a clue missing from real requests. They passed, but the rule failed in real use. I went back to the actual inputs to correct it.
I like having several projects on the go. In a mission, I channel that energy into the priorities agreed with the team and share what they need to carry on.
Meet me through my talks and interviews →
Start a conversation