Physics PhD with 10+ years building data pipelines and ML models. Currently seeking Data Scientist roles.
- Build end-to-end ML pipelines: ETL β feature engineering β modeling β evaluation β 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 Β· 42 analytical features Β· 108 tests Β· 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 β Character Traits Evaluator | End-to-end ML pipeline: 72 days behavioral data β NLP/sentiment analysis β 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 β Josh Starmer
- Think Like a Data Scientist β Brian Godsey
