I am a Generative AI Engineer and Software Developer with 3 years of experience building production-scale AI and backend systems at Amazon and Vibrium AI. Specialized in LLM applications, RAG pipelines, multi-agent AI systems, and AI voice automation. Experienced in designing scalable cloud-native solutions using Python, AWS, FastAPI, and modern AI frameworks, with a strong foundation in distributed systems, observability, and system reliability. M.Tech graduate from IIT Jammu passionate about transforming cutting-edge AI research into real-world products. I invite you to schedule a meeting with me to discuss how we can bring your vision to life!

• Engineered real-time AI voice automation infrastructure supporting 1L+ daily call sessions using AWS Bedrock, OpenSearch, Aurora RDS, and event-driven architectures.
• Developed a UI-driven multi-agent framework with dynamic agent-to-agent/human transfers, contextual memory preservation, and custom function tooling.
• Optimized Retrieval-Augmented Generation (RAG) pipelines using AWS Knowledge Bases and Titan embeddings, achieving <= 40 ms retrieval latency.

• Rebuilt a mission-critical internal log analysis platform, restoring 100% core functionality and improving system stability.
• Engineered and deployed an ETL-based LMK reporting system using AWS Glue, Athena, and S3, improving log traceability by 3x.
• Reduced logging noise by approximately 60% and improved debugging efficiency by 30% through enhancements in the PerfTracker architecture.

• Developed full-stack applications using Python, React.js, REST APIs, and MySQL, contributing to the delivery of scalable production-ready software solutions.
• Designed and implemented backend APIs and business logic modules, improving application responsiveness and reducing manual operational effort through workflow automation.
• Resolved 20+ software defects through debugging, root-cause analysis, and performance optimization, enhancing system reliability and user satisfaction.

• Agentic Voice AI System: Built a production-grade multi-agent voice AI system on LiveKit Cloud with intelligent routing, real-time STT/TTS (Cartesia, Silero), and function-calling for automated appointment booking.
• MyTripMate (Multi-Agent Travel Planner): Designed an adaptive travel itinerary engine using Google ADK and Vertex AI, featuring specialized sub-agents for logistics, booking, and dynamic real-time adjustments based on weather and delays.
• Insurance Document RAG Chatbot: Built a RAG chatbot with 90%+ retrieval accuracy using FAISS vector search, LangChain, and Groq LLM API, ensuring memory-aware context handling under strict token limits.

Build RAG App

Develop GenAI Applications

Google Data Analytics

Python for AI