// 18+ years — production GenAI systems, RAG pipelines, and multi-agent LLM orchestration (LangChain, LangGraph, LlamaIndex, CrewAI). Full-stack delivery with React, Angular, Next.js, Python, FastAPI — finance and legal domains.
Senior AI/ML Engineer and Full-Stack Developer with 18+ years specializing in production GenAI systems, Retrieval-Augmented Generation (RAG) pipelines, and multi-agent LLM orchestration (LangChain, LangGraph, LlamaIndex, CrewAI) across finance and legal domains.
Architected production GenAI systems including hybrid-search RAG platforms, multi-agent contract analysis workflows, and PII redaction pipelines using Python, FastAPI, and vector databases with leading LLMs (OpenAI, Anthropic Claude, Gemini, Llama).
Bridges GenAI engineering with enterprise full-stack expertise (React.js, Angular, Next.js, FastAPI) — shipping end-to-end AI-powered applications with production-grade UX, WCAG 2.1-compliant SPAs backed by robust CI/CD, comprehensive testing, and Agile delivery.
Senior Full-Stack Developer & AI Integration Lead — Angular / React / TypeScript
Citi Retail Services — internet banking platform with RxJS, NgRx, AEM, REST/GraphQL, WCAG 2.1 AA, enterprise security, and GitHub Copilot skills/agents for automated Jira-to-PR engineering workflows.
Gen AI Engineer / Full-Stack Developer — Internal Products
LangGraph multi-agent contract analysis, enterprise LlamaIndex RAG with hybrid search, and PII redaction pipeline — FastAPI, Milvus/Pinecone, Celery, RabbitMQ.
Project Lead
End-to-end delivery of mobile and handheld applications for US clients — retail and housing inspection apps with GPS, camera APIs, and offline sync.
CITI · full-stack & AI integration · 2014–present
Retail internet banking platform: Angular/React apps with RxJS, NgRx, REST/GraphQL and AEM integration, WCAG 2.1 AA, 60+ reusable components, enterprise security, and GitHub Copilot skills/agents automating Jira-to-PR workflows.
Gen AI · LangGraph multi-agent
LangGraph multi-agent platform for legal teams — 70% faster contract analysis with 95%+ clause identification accuracy, dynamic playbooks, embedding-based recommendations, OCR, and Milvus/Pinecone semantic search.
Gen AI · document privacy
End-to-end PII redaction pipeline (FastAPI + DeepSeek OCR) with React frontend — drag-and-drop upload, real-time preview, template-driven profiles, and GenAI entity recognition beyond regex.
LlamaIndex · hybrid retrieval
Production LlamaIndex RAG with dense + BM25 hybrid search (+35% retrieval accuracy), multi-modal ingestion for 12+ formats, Milvus at 500K+ embeddings, RabbitMQ indexing, and streaming citation API.
EOF — all systems nominal