Welcome to my GitHub profile! I am Enrique Barrueco, a passionate Data Scientist deeply interested in data-driven decision-making, predictive modeling, and crypto. My journey in data science has been shaped by a combination of strong academic foundations, practical experience, and a relentless curiosity to explore new technologies.
PunkPredictor.xyz is a subscription site that gives its users access to the most accurate CryptoPunk price prediction model. It leverages a deep neural network with an in-production median absolute percent error of 6.89%. Users also gain access to valuation intervals generated using conformal predictions and valuation reports created using explainable AI.
DFKData.io is a detailed dashboard embedded in a site that provides price and supply analytics for DeFi Kingdoms, a GameFi project. The data is directly queried from the DFK and Kaia blockchains, GraphQL, and various APIs.
We developed a ZK-verifiable model to predict the fair price of CryptoPunks, achieving an impressive 87% accuracy (1-MAPE). This project won the first prize for the best AI action, securing $3.5k.
At ETHDAM, I contributed to building a model that uses newly minted ERC20 contract data to predict market cap success. This project won the Slither Track and earned a $1K prize.
In this Kaggle-style competition, our team predicted the default of over-collateralized crypto loans, securing the 3rd prize with a $750 reward.
We developed an innovative app for fully private melanoma consultations using a picture. The neural network was hosted on the Nillion Testnet, and the app won the first prize in the Best Blind AI App category, along with $3k.
During my time working as a student assistant for Maastricht University, I contributed to two significant research projects, resulting in published papers:
-
Dark Web Market Scraping: Scraped 12 Dark Web markets to identify compromised student accounts of the university. Read the paper here.
-
Crypto Donations Chrome Extension: Participated in the development of a JavaScript Chrome extension that facilitates crypto donations to content creators. Read the paper here.
-
Master of Science in Data Science (Maastricht University, 2020 - 2024)
- Thesis: Predicting Cardiovascular Disease in Chronic Myelogenous Leukemia Patients
- Notable Projects: NLP on heraldic texts, NLP on YouTube videos for Bitcoin price prediction, Swarm bullying detection, E-commerce item pricing.
-
Bachelor of Science in Economics (Universidad Complutense de Madrid, 2014 - 2019)
- Thesis: Economic Analysis of Bitcoin
- Programming: Python (Pandas, NumPy, Scikit-learn, PyTorch, TensorFlow, HuggingFace, Transformers, Optuna, SHAP, Snorkel, spaCy, Selenium, Playwright, Matplotlib, Seaborn, Plotly), SQL, Dune SQL, GraphQL, SPARQL, JavaScript, TypeScript, HTML, CSS
- Data Science & Analysis: Power BI, Dune SQL, LIME, Spark, Predictive Modeling, Deep Learning, Natural Language Processing (NLP)
- Web Development: Selenium, Playwright, Microsoft Dynamics 365 (CRM and Business Central)
- Blockchain & Crypto: Web3, Dune Analytics
- Cloud Platforms: AWS, Vercel, Heroku, DigitalOcean, Microsoft Dynamics 365
- Soft Skills: Curiosity, Perseverance, Consistency, Resourcefulness, Clear Communication, Structure and Design of Data Problems
- LinkedIn: Enrique Barrueco
- GitHub: github.com/ennriqe
- Twitter/X @BarruecoEnrique
I'm always open to discussing new ideas, collaborations, or anything related to data science and crypto. Feel free to reach out!