FRANÇOIS AMAT // PhD · CTO · BUILDER

Research brain.
Builder hands

I’m François. I build full-stack products, design auditable AI systems, and turn research into practical tools.

1# Research → code → real-world impact.
2 
3profile = {
4    "name": "François Amat",
5    "background": "PhD, Explainable AI",
6    "experience": ["CTO", "Engineer"],
7    "stack": ["Python", "React"],
8    "focus":
9        "forward_deployed_engineering",
10}
11 
12# Curious? Open the terminal.
13print(profile["focus"])

portfolio shellPython UTF-8

François Amat / human behind the code
PhD in Explainable AI · Institut Polytechnique de Paris

while (curious) { learn(); build(); repeat(); }Scroll to explore

/* EXPERIENCE */Institut Polytechnique de ParisPhD · Explainable AIsuzanCTO · AI governancefingreen.aiCTO · 2022–2023Dassault Systèmes

01 / SELECTED_WORK

Less talk. More shipped.

From enterprise constraints
to research questions. Hands on, end to end.

CTO · AI governance

Making enterprise AI accountable.

Secure AI pipelines, policy oversight, and traceable architecture for an enterprise governance platform.

Enterprise AIArchitectureGovernance

CTO · 2022–2023

Turning carbon data into decisions.

Built the technical foundation and led engineering for a carbon analytics and ESG reporting platform.

PythonReactProduct ownership

PhD research · Explainable AI

Making relational data explainable.

A multi-table AI toolkit for exploring relational schemas and generating interpretable rules with domain experts.

Explainable AIRule miningOpen source

02 / README.md

Human context.
Systems thinking.

I like working where the problem is still taking shape. My path has taken me from explainable AI research at Institut Polytechnique de Paris to building and leading startup engineering teams.

Forward Deployed Engineering is where I want to bring those experiences together: close to the people using the product, hands-on with the systems behind it.

Explore my PhD Press kit

Start with the problem.

Understand the workflow, the constraints, and what would actually make someone’s work better.

Own the whole build.

Connect data, backend, and interface. Make the architecture serve the product.

Make it understandable.

Build for transparency and traceability. A useful system should be one people can question and trust.

François AmatEngineer, researcher, maker.

03 / PROJECT_INDEX

~/projects[31]

Research tools, weekend experiments,
and ideas that turned into products.

Showing 6 of 31 projects

ProductsLive
DemoBro
Personalized product demo videos for sales teams. Record a walkthrough or direct a demo with an AI assistant, then adapt it to your audience.
- AI
- Product demos
- Sales
Try DemoBro
Products2016-2017
Pawee
Veterinary online appointment platform and digital services.
Instagram Startuponly
Products2022-2023
Fingreen
Carbon footprint analytics and ESG reporting platform. CTO until exit (September 2023).
- Python
- React
- Climate Tech
- ESG
Visit Site
Products2023-Current
Suzan
AI compliance & governance platform (RegTech) for enterprise adoption. CTO.
- AI
- Enterprise
- Architecture
- Governance
- Compliance
- RegTech
See Blog Visit Site
AI & research2025
MATILDA
Multi-table explainable AI toolkit for relational datasets.
- Explainable AI
- Rule Mining
GitHub Contact for details
AI & research2025-Ongoing
MAHILDA
High-level declarative assistant for interpretable rule mining.
- Explainable AI
- Rule Mining
GitHub Contact for details

04 / DEV_LOG

Notes from the rabbit hole.

All writing

22 Mar 2026

AI / Benchmark

BakeMark: Benchmarking AI-Generated Landing Pages

I built a benchmark comparing landing pages generated by four LLMs — Claude Haiku 4.5, GPT-5.4, Claude Sonnet 4.6, and Claude Opus 4.6 — across 47 design prompts for a fictional bakery.

2 min read 16 Mar 2026

AI / Video