Alonso Valdés

• Developed a SaaS platform for architecture firms that automates the review of municipal records in two phases: document admissibility and regulatory analysis.

• Implemented the frontend in React, TypeScript, and Supabase with authentication, record management, intelligent document upload, real-time features, admin panel, SII/cadastre maps, and PDF reports.

• Integrated AI workflows with Gemini and Edge Functions, including contextual chat with SSE streaming, retrieval over urban regulations and private documents, and operational support for the pilot.

• Technologies: React, TypeScript, Supabase, PostgreSQL, Gemini API, Edge Functions, Cloudflare, Docker, RAG, LLM embeddings.

• Development of an interactive web application with Dash and Flask for the visualization of large volumes of data, optimizing browser performance through advanced metadata management.

• Construction of an ETL pipeline for 1M+ Spanish public procurement records, integrating web scraping from 6 platforms (PLACE, Catalonia, Community of Madrid, Andalusia, Galicia, La Rioja) with coverage from 2017–2025.

• Optimization of data loading times, achieving a reduction in response times through the implementation of Redis cache. Integration of dynamic filters with modern design, improving user experience in terms of usability and fluidity. Technologies Python, Dash, Flask, Docker, MariaDB, ElasticSearch, FastAPI, Plotly, SQL, Playwright, GitHub, Redis, Pydantic, OpenAI API, Docling, Streamlit, Beautiful Soup, Numpy,

Pandas, n8n, Supabase, LLM.

• Construction of an actuarial microsimulation projecting the Chilean pension system's expenditure until 2100, based on over 80M monthly contribution records (10% sample) from the Pension Data Information System.

• Developed stochastic projection models to analyze the sustainability of the Pension Reserve Fund using Monte Carlo simulation.

• Evaluation of the impact of the Employment Protection Law due to COVID-19 using propensity score matching and probit to measure the relationship between employment coverage and pension contributions.

• Technologies: Python, Stata, Numpy, Pandas, scikit-learn, Matplotlib, Excel VBA, Statsmodels,

Beautiful Soup.