Physicist | Data science | AI & Machine learning | Energy analytics
Machine learning researcher with a physics background and doctoral research focused on data-driven modeling of energy systems. My work spans time-series analysis, anomaly detection, and unsupervised learning techniques, including UMAP, HDBSCAN, and LOF. I develop Python-based workflows that combine machine learning, large language models, and knowledge graphs for structured analysis and decision support. My research focuses on building scalable, interpretable, and explainable solutions for real-world scientific and engineering challenges.
- Machine learning for solar energy systems
- PV anomaly detection and performance analytics
- Thermal modeling for outdoor PV operation
- Forecasting, data quality, and decision-support workflows
- Domain-specific AI tools with provenance and citations
| Project | What it shows | Links |
|---|---|---|
| Research Portfolio | A clean portfolio for PV, ML, and data science work. |   |
| Trustworthy Energy RAG Chatbot | Document-grounded technical-support chatbot with provenance, citations, OCR, and fault-code lookup. |  |
| SolarGraph AI | LLM-powered knowledge graph exploration for PV solar energy and materials-science research. |  |
| PV Power Predictor | Machine-learning workflow for predicting photovoltaic power output. |  |
Python | scikit-learn | FastAPI | Flask | Streamlit | SQL | MATLAB | Linux | Git | MLFlow
I develop research-oriented prototypes that connect solar-energy domain knowledge with machine learning, knowledge graphs, explainable analytics, and practical interfaces for scientific insight.
