My background is multicolor:
Studied criminalistics; worked for the Argentinean government for a couple of years; left the country to go travelling and started working at events and gastronomy; moved to Berlin where I managed a restaurant for 5 years... That's when I fully understood the power of data driven decision making!
🎓 I recently graduated from a Data Science & Machine Learning bootcamp at Code Academy Berlin, where I built projects applying machine learning, data wrangling, and visualization techniques.
🔍 I'm passionate about using data to solve real-world problems — especially in sustainability, public health, and urban mobility.
📂 My portfolio includes work with:
- 🏙️ Geospatial analysis (Capital Bikeshare DC & Public Transport proximity)
- 🌱 Sustainable construction (Predicting pollutants in building materials)
- 📰 Web scraping & NLP (Crypto news sentiment analysis)
- 📊 Dashboard development (Streamlit apps for financial/air quality monitoring)
Languages: Python, SQL
Libraries: Pandas, Scikit-learn, Matplotlib, Plotly, Folium, TextBlob, BeautifulSoup
Tools: Git, GitHub, Jupyter, VS Code, Streamlit, AWS Lambda, PostgreSQL, Railway
Other: Docker, Bash, Tableau, Canva
Here are some projects I’m most proud of (click to explore):
| Project | Description | Tags |
|---|---|---|
| 🚴 Capital Bikeshare Demand | Predicting ride demand based on station proximity to public transport, city center, landmarks and population density | Geospatial · Regression · EDA |
| 🏗️ Pollutant Predictor for Construction Materials | Bootstrapped a model to classify pollutants based on Ökobaudat data | ML · Classification · NLP |
| 📰 Crypto Sentiment Pipeline | Scraped financial news & analyzed crypto market sentiment with TextBlob. | Web Scraping · NLP · SQL |
| 🌍 Air Risk Index for Asthma | Aggregated NO₂, O₃, humidity and temperature into a spatial health index. | AQI · Satellite Data · Folium |
- Data engineering.. any tips? Not as glamorous as data science, but the base of it all ;)
- 💼 [LinkedIn] www.linkedin.com/in/pabs-soriano
- 💌 Email: pablosoriano1989@gmail.com
Thanks for stopping by! ⭐️