About me
Sixteen years in finance.
One conviction: AI agents change everything.
I've spent my career inside financial markets: algorithmic trading, machine learning applied to finance, leading data science and business intelligence teams. That background shapes how I teach AI agents: rigorously, concretely, and without hype. Finance is where my examples come from, not where the skills stop. What I teach is agent engineering, and it applies in any industry.
Hanane Dupouy
Author, trainer, and speaker on AI agents in finance, built on 16+ years in financial markets.
I started as an engineer (Arts et Métiers ParisTech, Télécom Paris), added a postgraduate degree in financial techniques (ESSEC), and built my career in financial markets: as an algorithmic trader, applying machine learning to finance, and managing data science and business intelligence teams.
When large language models arrived, I went deep: writing a book on building AI agents for finance, creating courses, training teams at international groups, and speaking at international conferences. Teaching finance professionals, engineers, and executives how AI agents really work is now my passion and my craft.
Credentials
Why people trust this work
📘 Author
Co-author of Building AI Agents for Finance (Packt): design and deploy financial AI agents with robust architectures, advanced reasoning, and Python. Also an O'Reilly video course, AI Agents in Finance, and three white papers on generative AI in finance.
🏦 Finance background
16+ years in financial markets: algorithmic trading, machine learning applied to finance, and managing data science and business intelligence teams. I know the constraints of this industry from the inside.
🏭 Corporate trainer
Designed and delivered in-company AI-agent programs for international groups: a 2-day technical intensive with hands-on labs for R&D engineers, and executive sessions helping a leadership committee shape its AI strategy roadmap.
🎤 Speaker & educator
Speaker at international conferences on AI agents in finance, masterclasses for private programs, and the three-level Mastering AI Agents in Finance curriculum.
Philosophy
How I teach
Raw first, then framework
You hand-build tool calling and the agent loop before touching any SDK, so no framework is ever a black box to you.
Finance-native
Labs are investment committees, earnings season, fundamental analysis. Finance is the setting, not the limit: what you learn is AI agent design first, and it transfers to any industry.
Reliability over demos
Guardrails, evaluation, observability, audit trails. An agent you can't defend in a risk committee doesn't count.
Start simple
Avoid agents by default. Increase complexity only when it's justified. Thirty seconds of honesty beats a month of hype.
Booking me to speak
Conference keynote or private masterclass: the formats, the signature talks, and where I've spoken are all on one page.
Let's talk
Training, speaking, coaching, or just a good conversation about agents in finance.