Vedang Mahajan

I’m an AI & Data Science undergraduate who enjoys turning ideas into production-ready systems—especially where strong backend design meets applied AI. I’ve built and deployed backend-focused solutions using Python and FastAPI, with an emphasis on scalable architectures from development through deployment. In my Council of AI project, I developed a multi-agent decision intelligence platform using LangGraph and integrated multiple LLM APIs for workflow orchestration. I engineered FastAPI REST APIs for state management and session persistence, and optimized memory retrieval with PostgreSQL to improve long-term conversational consistency. I also build end-to-end machine learning products. In the F1 Race Strategy Prediction Platform, I trained a hybrid LSTM model on F1 telemetry data and implemented a real-time strategy engine that recalculates pit windows dynamically; I designed dashboards to deliver live telemetry visualizations and strategy comparisons. Beyond projects, I’ve strengthened my technical depth through research and real-world building. I received an NVIDIA Academic Hardware Grant (32000 A100 GPU-hours) for JWST Deep Field reconstruction, working on diffusion and super-resolution models, and I authored a Scopus-indexed paper on a blockchain-based student management system using Hyperledger Fabric and smart contracts.