LinkedIn
I’m a AI/ML Engineer blending deterministic physics with large language models.
Currently researching AI Agents for Ground Penetrating Radar (GPR) simulations — a hybrid of physics-informed modeling and AI-driven data synthesis.
Current research:
- Building GPRSim, a retrieval-augmented and self-correcting simulation engine
- Developing multi-agent workflows using LangGraph + LangChain
| Project | Tech Stack | Achievement | Demo |
|---|---|---|---|
| EcoWise AI | Node.js, Angular, Gemini | 1st Place in MLH SharkHack'25 | View Project |
| AskWPI | RAG, LLM Finetuning, Agentic AI | SOTA Q&A in University Context | View Project |
| Jet Engine Remaining Useful Life | Python, PyTorch, Tensorflow | Kaggle Project | View Project |
| Gemma Finetuning on SQuaD | Python, Gemma3.1, HuggingFace, PyTorch | Model Finetuning | View Project |
| Accident Severity Prediction | Pandas, Scikit-Learn, PyTorch | Data Analysis & Prediction | View Project |
- Physics-informed neural networks (PINNs)
- Multi-agent RAG systems
- LoRA fine-tuning for scientific simulation datasets
Reach me at vijayrathank@gmail.com
Visit vijayrathan.com for my latest projects and publications
“Bridging physics and AI - one simulation at a time.”