I am a Software Developer at C3 AI, specializing in building enterprise-grade agentic AI systems. My work focuses on bridging the gap between complex data and actionable insights, specifically through Retrieval-Augmented Generation (RAG) and large-scale LLM integration.
Supply Chain Engineering Team - Order Allocation feature
•
Currently working on building an enterprise-grade agentic AI system that combines Retrieval-Augmented Generation (RAG) with OpenAI and Gemini LLMs to power context-aware, multi-step reasoning over large-scale enterprise data.
•
Designed and scaled distributed backend microservices in Python/TypeScript to power the C3 AI Supply Chain Suite, implementing multi-tenant RBAC, compliance logging, and workload isolation.
•
Maintained reliable operations of production Kubernetes inference workloads through autoscaling strategies and GPU scheduling optimization, reducing compute costs by 25% while sustaining latency below 200ms.
•
Implemented GraphQL APIs with TypeScript and Apollo Client for schema-driven type safety, eliminating REST endpoint duplication across forecasting and operational dashboards built in React.
•
Optimized React components using hooks (useReducer, useCallback, custom hooks), reducing code complexity by 20% and improving rendering performance across 15+ production modules supporting live operational views.
Built and deployed LSTM / DARTs / Keras forecasting models for energy microgrids; improved RMSE by 20% through hyperparameter tuning and feature engineering.
•
Automated MLOps lifecycle with Docker, Azure Pipelines, GitHub Actions, and model-drift monitoring; reduced manual retraining cycles by 35%.
•
Architected an energy optimization microservice that interfaced with IOT smart meters and DER devices through REST APIs, cutting energy costs by 22% for pilot households.
Worked for Charles River as an SDE Intern to develop a real-time market data handler using multithreading in Java to process data for equities and FIX trading.
•
Developed RESTful APIs in Spring Boot to onboard new clients into Charles River’s FIX system, integrating Azure AD SSO + RBAC policies, cutting onboarding time by 40%.
•
Created a React-based dashboard to monitor 10K+ FIX protocol agreements with real-time validation workflows and exception alerts, reducing manual verification time by 50%.
