Pablo Reyes

Reflexiones y tecnología. No tengo prisa.

Projects

X-RAI Vision

Portfolio auditing tool: upload a Morningstar PDF and the AI extracts the data, flags hidden risks/overlaps, and produces a written report plus a visual dashboard. Personal learning project to explore multimodal generative AI on financial documents, and to iterate on product UX end-to-end.
React Vite Tailwind CSS Node.js Express

OpenInsider

Automated pipeline that retrieves and parses insider trading disclosures from the Spanish CNMV, storing them in a database ready for analysis and visualization. Learning sandbox focused on resilient scraping, daily automation, and SQL-first persistence.
Python Scrapy PostgreSQL Docker

Barcelona Housing Analysis

Geospatial analysis of Barcelona’s housing stock since 1900 using Cadastre data, with maps and heatmaps to surface urban patterns over time. Learning project to combine data cleaning, GIS workflows, and exploratory visualization.
Python Pandas Matplotlib Pyproj Ipyleaflet

Carbitrage

End-to-end system that monitors marketplaces, processes listings in the cloud, estimates “fair” prices with models, and highlights outliers (potential arbitrage) in an interactive dashboard. Built as a learning project to practice data engineering on GCP, infrastructure-as-code, and ML for pricing.
Python Scrapy Pandas GCP (Pub/Sub, Cloud Functions, Firestore) Terraform

Backio

Interactive backtesting dashboard: load strategies, tweak parameters, and visualize equity curves, drawdowns, and metrics like Sharpe, CAGR, and more. Personal learning project to explore Backtrader and build an analysis UI with Dash/Plotly.
Python Dash Flask Backtrader Pandas

Sentiment Trading Strategy

Algorithmic trading strategy that blends sentiment analysis (news + social) to improve buy/sell signals, validated through backtesting. Created to learn how to combine NLP with quantitative evaluation and cloud deployment.
Python Backtrader NLTK (VADER) Google Cloud Firebase

AfterFacts

Static platform to track public predictions and compare them with real outcomes, creating a verifiable archive of “who said what, and when”. Experiment to learn Jekyll/GitHub Pages and design a simple, maintainable content workflow.
Jekyll Markdown GitHub Pages

La Liga Player Analysis

Segmentation of La Liga players using ML (PCA + clustering) to uncover statistical profiles and similarities beyond on-field positions. Learning project around scraping, unsupervised modeling, and visualization.
Python Pandas Scikit-learn Matplotlib Selenium

EntreTodos

Participatory democracy platform where communities can propose, discuss, and vote on initiatives with a clear lifecycle, reputation (karma), and search. Legacy project built as a learning exercise in Grails full-stack development and community/product patterns.
Grails Groovy Java Spring Security Bootstrap

Lyric Cloud

Web app that generates interactive word clouds from lyrics (or any text), with author search and dynamic visualization. Legacy project built to learn a full web stack and experiment with data visualization using D3.
Node.js Express MongoDB AngularJS D3.js

OCPJP6 Resumen Español

Spanish study notes for the Oracle Certified Professional Java SE 6 Programmer exam, distilled as a focused last-mile review resource. Personal study material published as a community resource.
Java OCPJP Documentación

Project image preview