Anuvrat Verma

AI-Driven Energy Supply Chain Resilience

– Engineered an autonomous 24/7 supply chain pipeline integrating an LLM-based risk analyzer with a custom C++ Min-Cost Max-Flow solver (SPFA) for real-time network rerouting – Architected a high-performance polyglot backend using gRPC and Protobuf to enforce strict, binary-fast contracts between C++, Python AI engines, and a React-Mapbox frontend – Designed an adaptive procurement framework leveraging parallel LangChain workflows, using HTTP/2 multiplexing to execute Tab 2 (economic impact) and Tab 3 (procurement simulation) streams concurrently – Developed a optimization engine for Strategic Petroleum Reserve drawdowns by blending LSTM-based Reinforcement Learning with Dynamic Programming solvers – Built a high-fidelity simulation sandbox supporting autonomous live-polling and dynamic "what-if" scenario testing, optimized via LCEL-driven token management