data scientist · barcelona

█

I build AI-powered products that turn complex data into decisions. Python, ML, LLMs, NLP — from raw data to production.

End-to-end builder: data collection, pipelines, ML models, agentic AI, and deployed applications. I ship things that work, not notebooks that sit.

interactive terminal — try a command or click one below
renzo@local:~$
$ cat pipeline.md
HOW I WORK
01 collect
Web scraping, APIs, enterprise data. I go get the data that answers the question.
02 transform
Pipelines that turn raw mess into analysis-ready tables. Pandas, SQL, validation at scale.
03 model
ML, NLP, deep learning, LLMs. Classical models and agentic AI — pick the right tool for the problem.
04 ship
Deployed apps, APIs, decision tools. Analysis that stays in a notebook helps nobody.
$ ls projects/
PROJECTS
~/projects/ds-radar
Agentic AI pipeline that scrapes job boards, extracts requirements with LLMs, scores listings against a structured profile, and ranks them. Multiple agents coordinate to automate a multi-step decision process into a single CLI command.
agentic ai · orchestration · decision automation
pythonai-agentsllmsrest-apis
source ↗
~/projects/no-botes-tu-voto
Decision-support tool for Colombian elections. Scraped policy positions, structured unstructured text with LLMs, built a scoring engine that matches voters to candidates based on policy alignment.
nlp · scoring model · deployed product
pythonllmsweb-scrapingprompt-engineering
live ↗source ↗
~/projects/un-speeches
Applied NLP and deep learning to 50+ years of UN General Assembly speeches. Topic modelling, sentiment analysis, and generative AI to surface patterns in diplomatic language over time.
deep learning · text analytics · topic modelling
pythontensorflownlpgcp
live ↗source ↗
~/projects/the-london-bible
Geospatial analysis of London neighborhoods. Ingested and transformed public datasets, built interactive maps with custom layers for transport, culture, and livability scoring.
geospatial · etl · data visualization
typescriptnext.jsmaplibresupabase
live ↗source ↗
~/projects/legalize-co
Open-source data pipeline that parses Colombian legislation from raw PDFs into structured, searchable data. Contributed extraction logic and data quality validation.
open-source · data engineering · etl
pythondata-pipelinesgitrest-apis
source ↗
~/projects/bjj-universe
Scraped and modelled the competitive BJJ graph. Built a 3D force-directed network visualization to explore athlete connections, rankings, and match history.
network analysis · 3d visualization · scraping
javascriptthree.jsd3.js
live ↗source ↗
$ ls ~/.skills/
SKILLS
llmsai-agentsnlpdeep-learningtensorflowscikit-learnprompt-engineeringpythonsqljavascripttypescriptbashpandasdata-pipelinesweb-scrapinggcpsupabased3.jsthree.jsmaplibrestreamlitmatplotlibdockergitlinuxvercelrest-apisnext.jscli
$ cat about.md
ABOUT
nameRenzo Rico
roleData Scientist
stackPython · LLMs · ML · NLP · SQL
locationBarcelona

Trained as an architect. Ended up building with data. Both fields reward the same thing: thinking in systems, obsessing over detail, and knowing when something isn't finished yet.

I want to build AI products that help people make better decisions — with a team that ships fast and cares about impact.

$ ./contact.sh
CONTACT