FilippoCappa

Economics and finance student in Turin, aiming for investment banking. I build things to test how I reason: a training engine for runners, a deal valuation model, and my own assistant.

Studying
BSc Economics and Finance with Data Science, University of Turin. Taught in English, graduating in late 2028.
Building
Roen Labs, an adaptive training engine for road runners. Founder since September 2026, in private alpha.
Aiming for
Investment banking, then a master's in finance.
Email meLinkedInGitHub

Every figure on this site has a small blue number next to it. Select it to see where the figure comes from.

Work

One company, one piece of finance work, one tool I built for myself. Each started from a question I wanted answered properly.

Live, private alpha

Roen Labs

Role
Founder, self-employed. Designed and built it alone in Python.
When
September 2026 to now
Built with
Python, FastAPI, Postgres (Supabase), Next.js, Modal, Strava API

An adaptive training engine for road runners, from 5 km to the marathon. It checks whether a goal time is realistic, builds the plan to race day, and re-plans after every run, explaining each change.

Stage
Private alpha, about 15 athletes1
Product
Engine, API and web app, all live2
Decisions
Rule-based and reproducible. AI only reads free text
Regression check
954 plans regenerated on every rule change3
Read about Roen Labsroenlabs.com

Sources

  1. 1Founder's count of invited athletes, October 2026.
  2. 2Live product.
  3. 3tools/sweep_event_specs.py and the engine's mathematical specification. The repository is private.

In development

Deal Autopsy

Role
Solo project
When
October 2026 to now
Built with
Python, pandas, SciPy, Excel

A Python project reconstructing the valuation of a real M&A deal using only the information available at the announcement.

Methods
Trading comparables, precedent transactions, DCF, LBO1
Built so far
The look-ahead guard and its tests2
Deal
Being selected1
Valuation summaryImplied value per share. Ranges appear once a deal is chosen.
Trading comparables
Multiples of listed peers
Not computed yet
Precedent transactions
Multiples paid in similar deals
Not computed yet
Discounted cash flow
WACC from CAPM, regression beta
Not computed yet
Leveraged buyout
Highest price at a target IRR
Not computed yet
Offer price, marked here once the deal is chosen
Read about Deal Autopsy

Sources

  1. 1Project brief and architecture notes, October 2026.
  2. 2src/deal_autopsy/data/asof.py and tests/test_asof.py. The repository is private until the first deal is published.

In personal use

Aide

Role
Built for myself
When
October 2026 to now
Built with
Python, FastAPI, Claude API, SwiftUI, Whisper, Tailscale

My own assistant. It runs on a computer at home, and I talk to it from an iPhone app I wrote. It handles my calendar, email, health data and university courses, and asks before doing anything that matters.

Runs on
A desktop at home, with an iPhone app1
Specialists
Secretary, professor, coach, finance1
Safety
Nothing is sent or changed until I confirm1
Network
Private. Not reachable from the public internet1
Read about AideSource on GitHub

Sources

  1. 1Project README.

About

What draws me to finance is how valuation turns uncertain information into a number you can defend. I like testing that kind of reasoning with data and code rather than only reading about it, so most of what I build has the same shape: explicit assumptions, rules that can be checked, and results that can be reproduced.

I'm in the second year of an English-taught degree in Economics and Finance with Data Science at the University of Turin, and I plan to go on to a master's in finance. Before university I studied applied sciences at Liceo Gramsci in Ivrea.

Outside class I run, and Roen Labs started there. I wanted a training plan that reacted to how my weeks actually went and couldn't find one I trusted, so I built it.

What I work with

  • Python
  • pandas and NumPy
  • SQL (Postgres)
  • Excel
  • FastAPI
  • TypeScript and Next.js
  • Swift