• Built an LLM-powered, plug-and-play Chatbot as a Service platform with retrieval-augmented generation (RAG), prompt retrieval, and real-time prompt updating, enabling teams to launch and customize chatbots without engineering overhead.

• Developed a scalable Knowledge Base Management System with flexible schema and data isolation that supports upserting, deleting, listing, and semantic searching of KB content using vector embeddings generated by OpenAI and other providers.

• Implemented a re-embedding API within the Knowledge Base Management System to enable rapid experimentation and seamless migration of KB content across embedding providers while maintaining retrieval accuracy.

• Developed scripts to ingest KB content from Excel and bulk-insert it into the system via APIs, streamlining onboarding for partner teams.

• Added observability metrics for latency, error rate, throughput, and availability to ensure reliable, production-grade performance.