Madrid · Open to quant, fintech and AI engineering roles
I build automated trading systems and operate them live.
For the past year I have run an automated market maker on the order book of Polymarket, a prediction-market exchange, quoting index, commodity and event markets from my own pricing models. It has made $39,866 net on $4.95M of notional, most of it from liquidity-provider rewards.
Before working for myself I spent three years as an investment analyst at Merlin Properties, an IBEX 35 listed real estate group. I build the models, the systems that trade on them, and the operations that keep them running unattended.
Live system, public record
- Liquidity rewards
- $43,480
- Rebates & yield
- $2,141
- Trading P&L
- −$4,072
- Fees
- −$1,683
- Net result
- +$39,866
$4.95M notional (face value) · $2.88M traded in cash · 2,924 markets
Figures as of 2026-10-12; refreshed automatically and cross-checked against Polymarket's public API.
01Flagship
Polymarket market making
The venue pays for providing liquidity, not for being right. The system prices each market from its own models, decides what risk it is willing to hold, and quotes both sides only where the numbers justify it — then stands down when they do not.
- Prices markets independently rather than trusting the venue's own midpoint
- Quotes only within limits it sets for itself, and declines when the edge is not there
- Runs unattended, with risk controls that stop it rather than let it trade on bad data
- Python
- asyncio
- IBKR TWS API
- WebSocket
- Polygon PoS
- Docker
- systemd
- Pyth Network
02Selected work
Systems I have built
Balobrador ordering and payments
LiveA full-stack ordering system for a cheesecake shop near Madrid: catalogue, cart, pickup scheduling and Stripe card checkout, with every price set on the server rather than in the browser.
- Oct 2025Live since
- StripeCard payments
- Stripe
- Astro
- React
- TypeScript
- +4
Options payoff visualiser
LiveSixteen option structures with payoff diagrams that recompute as you move a single variable. Built to sharpen my own intuition for how strike, premium and the underlying's price shape a position at expiry.
- 16Structures
- EN · ESInterface
- Next.js
- React
- TypeScript
- Tailwind
Market scanner and content pipeline
LiveAgents that watch prediction markets continuously for pricing anomalies, plus an unattended pipeline that turns a written script into a finished narrated video without a person touching the timeline.
- 24/7Coverage
- 0Manual steps
- Python
- Astro
- ElevenLabs
- Whisper
- +2
Seasonality and relative-performance dashboard
Hosted demoA research dashboard over 47 assets running seasonal-trend decomposition with configurable significance thresholds, plus multi-asset relative-performance comparison. Updates daily.
- 47Assets
- DailyRefresh
- Python
- Streamlit
- statsmodels
- Plotly
- +1
03What I do
Four things, one skill set
The same discipline at different altitudes: understand the mechanism, model it honestly, build something that survives being left alone.
Trading systems & microstructure
Order-book mechanics, quoting and risk logic, exchange APIs, latency-aware async design, and the operational scaffolding that keeps a system running with nobody watching.
Polymarket market maker — live since September 2025
Data, models & fair value
Pricing models, volatility, and multi-source data pipelines that fail closed rather than quietly serving stale numbers.
The market maker's fair-value engine, and the options visualiser
AI & automation engineering
Agents that monitor markets continuously, unattended media pipelines, and the plumbing that turns a manual process into one that runs itself and says when it breaks.
24/7 market scanner and automated video pipeline
Full-stack web and payments
Production web applications end to end: static-first front ends, serverless back ends, and card checkouts where the server — never the browser — decides what a customer pays.
Ordering and Stripe checkout at balobrador.es — live since October 2025
04Background
Finance first, then the engineering
I learned markets before I learned to ship software, which is why the systems I build are shaped by what actually goes wrong in one. Three years inside a listed REIT taught me how institutions handle reporting, risk and capital. Working for myself since 2022 taught me the rest, because when it is your own money and your own infrastructure there is nobody to hand the problem to.
Full CVStack
- Python
- asyncio
- pandas
- NumPy
- TypeScript
- React
- Next.js
- PostgreSQL
- WebSockets
- Docker
- systemd
- Linux
- Git
- IBKR API
- Stripe