Julian Hoang

Led development of an enterprise-scale Retrieval-Augmented Generation (RAG) platform for domain-specific conversational AI. Designed end-to-end ML pipelines, integrated semantic search and metadata tagging, and tuned production alarms for hallucination reduction. Drove low-latency inference and governance for 100M embeddings and OSS integration. Collaborated with cross-functional teams to ship dashboards and monitoring capabilities for enterprise use.

Enterprise-scale RAG platform for domain-specific conversational AI and operational teams, integrating semantic chunking and metadata tagging, achieving low-latency lookups across 100M embeddings. Focused on improving factual accuracy, retraining workflows, and governance for production-grade AI services.

Led enterprise-scale RAG platform for domain-specific conversational AI and operational teams; integrated semantic chunking, metadata tagging, and keyword-based search; enabled low-latency lookups across 100M embeddings; tuned models on domain data; integrated OpenAI GPT on AWS with monitoring dashboards; reduced hallucinations and achieved company-wide adoption of the system.