Data Scientist with 1+ years of experience building end-to-end ML systems — from raw data pipelines to containerized, production-ready applications.
I work at the intersection of classical ML, NLP, and GenAI, with a focus on shipping things that actually run in production, not just in notebooks.
- ML systems end-to-end — data ingestion, feature engineering, model training, serving
- Semantic search & RAG pipelines — vector databases, embeddings, retrieval
- Containerized applications — Docker, docker-compose, reproducible environments
- Cloud-deployed models — Azure ML, scalable inference
Generates personalized tourist routes in Spain from a natural language description.
- Embeds user input with a multilingual sentence transformer
- Queries Qdrant (vector DB) with geospatial filtering over 10k+ Spanish POIs
- Optimizes the route with Valhalla routing engine (car / walking / bicycle)
- Renders an interactive map + auto-generates a PDF brochure with Wikipedia images
- Fully containerized with Docker Compose — runs locally with one command
ML & Data: Scikit-learn · PyTorch · Pandas · Sentence Transformers · Qdrant
MLOps & Infra: Docker · Azure · pre-commit · pytest
Languages: Python · SQL
- Building production ML pipelines with proper MLOps practices
- Generative AI — LLMs, RAG systems, and AI agents
- Deepening cloud architecture knowledge (Azure)
