Senior @ UC Berkeley • Data Science + Cognitive Science
Interested in ML, data systems, and building things that actually get used!

• Building ML pipelines to parse and structure unstructured data (emails, PDFs, etc.)
• Experimenting with LLM-based systems and agent-style workflows
• Working on predictive models using real-world datasets (like F1 race data)

AI Email Processing Pipeline
Built a system to extract structured data from messy logistics emails using LLMs + Postgres. Focused on making the pipeline reliable and usable, not just a demo.

Minecraft Agent Evaluation Framework
Worked on evaluating AI agents in a simulated environment, including tracking progress toward goals and designing better scoring methods.

F1 Race Prediction
Used historical race data to engineer features and train models to predict race outcomes. Focused on understanding what actually matters (driver form, track history, etc.)

Languages: Python, SQL, JavaScript
ML/Data: Pandas, scikit-learn, Matplotlib
Tools: Git, Jupyter, Postgres
Web: React, Next.js

• ML engineering
• Data pipelines + backend systems
• NLP / LLM applications
• Real-world ML (not just notebooks)

Motorsports (F1), baking, mixing music, and random side projects

• LinkedIn: https://www.linkedin.com/in/shreyasingh3/
• GitHub: https://github.com/shreyasingh3