MS Data Science @ UMD. I build production ML systems — cloud-native pipelines, fine-tuned diffusion, heterogeneous graph models, and real-time inference services.
- Shipping: cloud-native ML pipelines on AWS + GPU inference services
- Working in: PyTorch, Diffusion, GNNs, MLOps, streaming data
- Learning: distributed training, LLM fine-tuning, model serving at scale
- Reach me: evivek@umd.edu
↑ each icon links to the repos in my profile that use that tech.
|
|
|
|
|
|
| Meshwatch Graph-native fraud detection in real time. Heterogeneous GNN (PyTorch Geometric) + XGBoost ● Python ★ 1 · pushed 6mo ago | ClearShot A diffusion-based pipeline for enhancing amateur product photos to professional quality. U ● Python ★ 1 · pushed 6mo ago | MedQA A modular NLP system that classifies medical questions, extracts symptoms, retrieves relev ● Python ★ 1 · pushed 7mo ago |
|
|
|
↑ On repeat — click the card to listen on Apple Music.
Thanks for stopping by — drop me a line at evivek@umd.edu.


