Hang Z.

Senior full-stack engineer with 10 years across Java/Go/Python distributed backends — and since 2024, building production AI/LLM systems. I help teams move LLM demos into scalable, observable agentic products. Currently building Morfeus, a multi-agent FX market intelligence platform that orchestrates 33 specialized agents into consensus-backed research. Previously at Cheetah Mobile: shipped a 40M-call/day Netty data service and a 10TB/day streaming pipeline that cut decision latency to seconds and infra cost by 30%. Stack: • AI/LLM: OpenAI, Claude, Gemini, multi-agent orchestration, RAG, tool calling, LangChain/LlamaIndex, vector DBs (Pinecone, Weaviate, Qdrant) • Backend: Java/Spring Boot, Go, Python/FastAPI, Netty, microservices, Kafka, Redis • Frontend: React, Next.js, TypeScript • Data & Infra: PostgreSQL/Supabase, MongoDB, Elasticsearch, AWS (Lambda / S3 / Redshift), Docker, CI/CD What I deliver: • Production-grade LLM apps — agents, RAG, evaluation, observability — not prototypes • High-throughput backends that survive real load (10TB/day, 40M calls/day track record) • End-to-end ownership: design → ship → monitor Open to long-term retainers and well-scoped projects. Let's talk about what you're building.