● Led end-to-end delivery of AI features powered by LLMs across multiple enterprise systems and userfacing platforms, increasing operational efficiency by 35%. ● Designed and built Retrieval-Augmented Generation (RAG) pipelines using LangChain + pgvector for internal knowledge Q&A and semantic support agents, improving answer accuracy by 42%. ● Implemented OpenAI Function Calling agents to automate product recommendations, onboarding, and workflow decisions, reducing manual effort by 50%.

● Spearheaded the company’s first LLM initiative — from infrastructure design to capability prototyping — establishing a foundation for 100% of future AI-driven features.● Built early prompting pipelines for classification, summarization, and natural language autofill using GPT-3.5, reducing manual processing time by 40%. ● Structured metadata, embedding schemas, and ETL pipelines to support vector-based semantic search, enabling a future 35% improvement in knowledge retrieval accuracy.

● Architected and delivered cloud infrastructure and DevOps systems to scale LLM workloads, secure inference keys, and support high-availability AI pipelines, achieving >99.99% uptime. ● Provisioned AWS Lambda, EC2, RDS, and S3 environments, enabling future-ready LLM inference endpoints and elastic scaling under load. ● Built CI/CD pipelines using GitHub Actions, Jenkins, and AWS CodePipeline, reducing deployment time by 60% and improving rollback reliability by 100% for AI services.