Physics PhD with 10+ years building data pipelines and ML models. Currently seeking Data Scientist roles.
- Build end-to-end ML pipelines: ETL → feature engineering → modeling → evaluation → dashboards
- Ship production-ready code: CI/CD, testing, documentation, containerization
- Communicate insights to technical and non-technical stakeholders
Soccer Physics Engine: A decision-support tool for soccer coaches and analysts. Takes player tracking data (22 players, 25fps) and produces tactical analysis, physical load monitoring, and movement intelligence.
Live Dashboard · 42 analytical features · 108 tests · Deployed on AWS ECS Fargate
Built with: Python, NumPy, SciPy, scikit-learn, PyTorch, FastAPI, Docker, AWS, React, D3.js
AiiDA-OLCAO: Python plugin automating HPC workflows with RabbitMQ, PostgreSQL, and SLURM integration
| Project | Description |
|---|---|
| Soccer Physics Engine | Tactical intelligence + load monitoring + player movement analytics from tracking data |
| CTE — Character Traits Evaluator | End-to-end ML pipeline: 72 days behavioral data → NLP/sentiment analysis → LLM-based trait profiles |
| Python LLM Playbook | Unified interface for OpenAI/Anthropic/Gemini/Groq/Ollama with consistent patterns and tests |
| NASA ADS Metadata Retriever | Automated pipeline to extract research paper metadata via REST API |
| Galactic Neighbors Finder | KD-tree based neighbor search for galaxy catalogs |
- Deo, D. K. "Are Recently Quenched Ellipticals Truly Isolated Centrals?" arXiv:2601.09846, 2026 (single-author)
- Deo, D. K., et al. "Investigating Quenching in RQE Galaxies with HI Studies" arXiv:2601.08027, 2026
LangChain, LlamaIndex, RAG pipelines, AI Agents, MCP
- The StatQuest Illustrated Guide to Neural Networks and AI — Josh Starmer
- Think Like a Data Scientist — Brian Godsey
