Featured Projects
Jan 2026 - Present
AI Physician Scheduling Agent
A healthcare AI assistant that turns physician scheduling preferences into structured, validated inputs for clinical operations.
Key Features
- Extracts availability, constraints, and preferences from natural-language conversations.
- Combines LangGraph orchestration, retrieval, validation guardrails, and human review.
- Designed for secure deployment into production healthcare scheduling workflows.
PythonLangGraphLangChainAzure OpenAIRAGPostgreSQLDockerKubernetesFastAPI
Code not public due to healthcare and partner confidentiality.
Apr 2026
arXivLens - AI Research Paper Assistant
An AI research assistant for faster literature discovery across arXiv papers using semantic search, reranking, and grounded answers.
Key Features
- Supports natural-language questions over scientific paper collections.
- Uses metadata filters, embeddings, and cross-encoder reranking to improve retrieval quality.
- Generates grounded answers from retrieved paper context instead of unsupported summaries.
PythonFastAPIStreamlitChromaDBSentence-TransformersOpenAI APICross-Encoder Reranker
Dec 2025
Telco Churn Prediction - End-to-End ML Project
A production-style machine learning system that predicts telecom customer churn and serves predictions through an API and web UI.
Key Features
- Trains an XGBoost churn classifier with feature engineering and experiment tracking in MLflow.
- Serves predictions through a FastAPI endpoint and Gradio interface for quick manual testing.
- Packages the app with Docker and supports repeatable deployment through GitHub Actions and AWS ECS Fargate.
PythonXGBoostFastAPIGradioMLflowDockerGitHub ActionsAWS ECS