Dear Gregoire,
I am writing this open letter because I am applying to the MSc Data & AI for Finance from Albert School x Mines Paris-PSL, 2026-2028 cohort. You probably receive hundreds of these. I am not trying to be one applicant among many. I am trying to convince you, in four minutes of reading, that I am already the profile this program is built for.
My name is Mathieu Haye, I am 22, based in Paris, and I just secured a two-year apprenticeship as ALM Pilot Apprentice at AFD (Agence Francaise de Developpement), starting September 2026. Rhythm: 4 days company, 1 day school. I am applying to Albert School with that apprenticeship already in hand, which means one concrete thing: you do not place me in a classroom, you place me on an ALM desk from day one.
I want to learn quantitative finance, not because it looks good on a CV, but because in a few months of freelance work since October 2025, I have built trading bots, backtests, investment pipelines, and every time I hit the same ceiling: I can ship, I cannot yet prove my numbers.
Why Albert School, specifically
Since Albert School was founded in 2022, you have defended a thesis that speaks to me: the academic rigor of Mines Paris-PSL, the business reflex, plus access to the industry. The partnership signed in December 2024 between Albert School and Mines Paris makes that promise concrete. It is the exact intersection I need. Not a finance MBA that skips the math. Not a quant M2 that forgets the business. An MSc where you learn to think like a quant and ship like a builder.
I hesitated between several programs. Three reasons I made Albert School my first choice:
- The pedagogical ambition. An MSc Data & AI for Finance co-built with Mines Paris is a strong bet. You are not copying an existing program, you are creating one. That matches my builder temperament.
- The human scale. A tight cohort, instructors who still practice finance or data the rest of the week. I would rather learn from a quant who trades than from a quant who teaches.
- The fit with my apprenticeship. AFD does ALM on a public-mandate balance sheet. Mines Paris has a real tradition in stochastic modelling. Albert School teaches me the business language of that risk. The three pieces fit perfectly.
What I bring to the cohort
I am not coming to Albert School to be trained. I am coming to get more precise about what I already do. Seven finance / data / AI projects are in production or piloting today. Here are three that summarise the rest:
A Bloomberg-like dashboard plugged into Claude
I track my multi-asset portfolio (BTC, ETH, SOL, NVDA, TTE, CW8) through a Next.js 14 dashboard I coded myself. It computes RSI, MACD, Bollinger, ATR across four timeframes, runs 10-day Monte Carlo simulations, computes Pearson 6x6 correlation of my holdings, and at 08:50, 13:00 and 17:30 Paris time calls Claude Haiku 4.5 to produce a market commentary delivered to my Telegram. Trade journal, DCA tracker, three-tier alerts. All this for a personal portfolio. Because I need to see what is happening, not just live with it.
A crypto mean-reversion bot on MEXC
RSI + Bollinger Bands on BTC and ETH. Paper trading 90 days: +3.2%. Live pilot 30 days: -2.1%. Taker fees and slippage ate the edge. I could have hidden it. I bolded it on my portfolio. Because the lesson matters more than the performance: a backtest that does not model friction costs tick by tick is a lie. That is exactly the rigor I expect from Mines.
An Ichimoku indicator backtested over 5 years
Pine Script v6 on TradingView, Python backtester on the other side to replicate the logic candle by candle, fees included. On AAOI over 5 years: +187%, 54% win rate, PF 1.62, MaxDD 24%. Buy & hold on the same window: +1088%. My indicator loses in absolute terms, but cuts the drawdown in half. It is not alpha, it is risk management. The difference is precisely what an ALM course at Mines Paris lets you formalise.
Why now
I built what I could on my own. It works. It pays clients. It convinced AFD. But I am hitting a limit: I can build, I cannot yet prove. The MSc Data & AI for Finance from Albert School x Mines Paris-PSL is the frame that turns an operator into a serious quant profile. Then I will be able to really build the decision tools I care about, on a trading desk, an asset manager, or a quant fintech, with the science behind them.
The AFD apprenticeship is the first stone. The MSc is the structure that holds it all together.
My ask
I am not asking for a favour, I am asking for four minutes of attention. You will find the rest of my profile on mathieuhaye.fr: the seven projects detailed, modals with screenshots, full timeline, contacts. The site is hand-coded (HTML, CSS, JS, no framework, no build). I would rather have my first demo of rigor visible from the URL.
If my application speaks to you, I am reachable at the contacts below. If it does not, or if there is something you want to challenge before admission, I am also up for a chat. I learn faster from one "no, here is why" than from ten "we will get back to you".
I would be proud to join the 2026-2028 Albert School cohort.