OVERVIEW
Principal Generative AI Engineer and Architect specializing in the design and deployment of production-grade GenAI systems, Advanced RAG pipelines, and Agentic workflows. Expert at bridging the gap between raw LLM capabilities and high-scale, low-latency production environments. Instrumental in building the Vooz ecosystem serving over 500k monthly active users. Committed to a "Systems First" philosophy—building reliable, observable, and cost-optimized AI solutions.
SKILLS
GENAI & AGENTS
LangGraph, AutoGen, CrewAI, LlamaIndex, OpenAI, Function Calling, Prompt Engineering
VECTOR DBS & RETRIEVAL
Qdrant, Pinecone, Milvus, Chroma, LanceDB, Advanced RAG, GraphRAG, Hybrid Search
AI INFRASTRUCTURE
Triton Inference Server, ONNX Runtime, vLLM, Ollama, Dynamic Batching, LoRA, QLoRA
DEVOPS & MLOPs
Azure AKS, ArgoCD, Azure Front Door, OpenTelemetry, Ragas, TruLens, DeepEval, Guardrails
FRONTEND & APPS
Next.js 16, TypeScript, React Native, WebRTC, Framer Motion
BACKEND & DATA
Node.js, Python, SignalR, Memgraph, PostgreSQL, Redis
EXPERIENCE
VOOZ INC
Principal GenAI Architect (Founding Engineer)
- Developed a dual-track moderation pipeline integrating backend LLMs with local edge-inference for real-time safety responses.
- Spearheaded Triton Inference cluster with Dynamic Batching, achieving 1280+ RPS with full OTel tracing.
- Engineered browser-local inference using ONNX Runtime (WebGPU) for real-time content processing.
- Engineered multi-tenant Next.js environment utilizing Parallel and Intercepting Routes, scaling to 500k+ MAU.
- Orchestrated multi-cloud deployments secured by ArgoCD and Azure Front Door.
AIMICA LTD
Software Developer (Lead Role - AI & RAG)
- Designed and implemented a scalable RAG-based knowledge system integrating open-source LLMs.
- Led a team of 3 developers building AI-powered mobile apps using React Native (Expo).
- Developed backend services (Node.js/Python) to support AI orchestration and caching.
INFOEDGE
Mobile Developer
- Optimized critical React Native flows for high-traffic apps, resolving memory leaks.
PROJECTS
Multi-Agent Workflow Orchestrator
Autonomous agent system using LangGraph and AutoGen for complex task planning and tool execution.
Medical Paper NLP Categoriser
Deep learning model using TensorFlow on PubMed datasets. Achieved 90%+ accuracy in classifying medical abstracts.
EDUCATION
B.Tech in Computer Science
Dr. A.P.J. Abdul Kalam Technical University