📈 Data Scientist & Analytics Engineer @ Primetag 🇵🇹
🎓 MSc Data Science — Universidade do Porto
🤖 Currently exploring: LLM engineering · RAG pipelines · Agentic AI
- Building and automating data pipelines that save real time and money
- Forecasting, causal inference, and geospatial analysis applied to business problems
- Deploying ML models end-to-end — not just training them
- Extracting value from unstructured data using Azure AI and LLMs
Languages: Python · R · SQL
ML & AI: scikit-learn · XGBoost · LangChain · ChromaDB · Streamlit
BI & Viz: Power BI · Tableau · R/Shiny
Cloud & MLOps: Azure (ML · Vision · LLM) · Docker · GitHub Actions
Other: dbt · Git · JIRA · Agile
| Project | Description | Stack |
|---|---|---|
| Household Finance Copilot | AI-powered personal finance platform: ingests bank statements via Gmail, extracts transactions with Groq/Llama 3, review queue, analytics dashboard — fully deployed | FastAPI · React · Supabase · Groq · Docker |
| GeoAI Airbnb Intelligence Platform | Geospatial ML platform explaining what drives Airbnb performance in Porto — spatial feature engineering, LightGBM, SHAP explainability, H3 hexagon maps, FastAPI backend | Python · LightGBM · SHAP · FastAPI · DuckDB · Deck.gl |
| TikTok Brand/Creator Classifier | Cost-sensitive ML pipeline separating brands from creators using profile signals: bios, usernames, emojis, engagement patterns | Python · XGBoost · scikit-learn |
| Influencer Research RAG | Local RAG assistant to query academic papers with cited answers — no API keys, runs on-device | Python · ChromaDB · Qwen2.5 · Streamlit |
| Data Analyst Job Skills | What skills and salaries actually look like for data analysts — scraped from LinkedIn | Python · Jupyter · GitHub Pages |
| Causal Effect of Layoffs on Stock Price | Event study + DiD to measure how layoff announcements move stock prices | R · causalimpact |
 |
 |
|---|---|
 |
 |
 |
 |
 |
 |
