Curious by default, quantitative by choice
I am Milan Killian, a recent graduate in Commercial Economics on my way into quantitative finance, building things with markets, data and AI as I go.
The short
version
Economist by training, builder by temperament. During my Commercial Economics degree I kept drifting toward the quantitative questions: why prices move, what separates signal from noise, and how you prove that something actually works instead of just claiming it.
So I started building: LLM classification pipelines, validated knowledge bases and agentic workflows for financial institutions. Everything on this site follows the same rule. Keep it deterministic where you can, and keep a human in the loop where it matters.
My edge is that I sit in the middle: I can talk to the business side, I write the code myself, and I know what it takes to get LLMs and agents working inside a real team’s processes. Financial institutions are where all of that lands for me.
The route
Graduated · Commercial Economics
Final thesis on how LLMs and AI agents can be put to work inside an organisation’s teams and processes.
Pre-master’s · Finance
Closing the gap to graduate-level finance: statistics, econometrics, asset pricing.
Duisenberg Honours Programme · Quantitative Finance
Master’s specialisation in quantitative finance: derivatives, risk, computational methods.
Quant & AI within financial institutions
Research discipline on one side, production systems on the other. That combination is what I am working toward.
How I work
Question
Every strategy starts as a doubt. Before I trust a model or my own code, I try to break it first.
Validate
Nothing ships without proof. If a claim is not backed by a test or a measurement, it counts as an anecdote.
Deploy
Ideas only count once they run. I build systems meant for production, with risk controls in from day one.
Ownership
From concept to kill switch, the whole chain is mine. If it breaks in production, that is on me too.
Compound
Every project feeds the next one. The code and the lessons carry over, and that stacking is where the real returns are.
Still curious
Markets podcasts, poker maths, and an unreasonable number of side projects. The mosaic M on the home page? Click the bubbles.
Let’s talk
Open to quant-adjacent roles, internships and good conversations about markets, data and AI.