Sree Sai Bhargav Sarvepalli

Full-stack SaaS platform with real-time collaborative Kanban boards, drag-and-drop with optimistic UI, and Flow — a visual git-style project map. Built Nex, an AI assistant (Anthropic API) that creates tasks via natural language, generates descriptions, and summarizes project activity. Multi-user workspaces with Supabase Realtime, Row Level Security, and Google OAuth.

Implemented a Quantum SVM using Qiskit's ZZFeatureMap and FidelityQuantumKernel to classify network intrusions on the KDD Cup 99 dataset. Benchmarked quantum kernel performance head-to-head against a classical RBF-SVM across accuracy, F1, and ROC-AUC. Built a React 18 + Three.js landing page and a Streamlit dashboard with precomputed cached results.

Cloud-deployed ML microservice using FastAPI + LightGBM to classify pulsar candidates from telescope features. Independent microservices for ingestion, preprocessing, inference, and storage. Streamlit dashboard with batch predictions, SHAP interpretability, and EDA visualizations deployed on AWS EC2.

Multilingual emotion-classification workflows using mBERT, XLM-R, and XLNet. Evaluated across multi-domain datasets with reproducible benchmarking pipelines and optimized tokenization strategies.

CNN models to classify chest X-rays into COVID-19, Pneumonia, and Normal categories. Applied data augmentation, EDA, and GPU-accelerated training to reduce overfitting on imbalanced medical imaging datasets.