AI Engineer (Multi-Agent LLM Applications)
An AI engineer who builds production LLM services: multi-agent chat orchestration with intent routing, RAG grounding, structured output and guardrails, delivered as FastAPI services on AWS. Also brings a computer-vision background, fine-tuning and ONNX-optimising recognition models for client deployments.
- Level
- Mid
- Experience
- 4–6 years
- Based in
- Vietnam
- Languages
- Vietnamese (native), English (intermediate)
Selected work
- Retail operations (Japanese enterprise client): implemented free-chat orchestration that routes store-manager questions to data-QA, root-cause, reporting and dialogue sub-agents, with RAG grounding and structured outputs to reduce hallucination.
- Enterprise knowledge management: built a document-QA assistant with ingestion, chunking, metadata filters and a guardrail that declines to answer when no relevant context is found.
- Business reporting: built an LLM service that turns semi-structured inputs into validated report sections and action items, with schema checks and automatic regeneration.
- Industrial metering (Japanese client): fine-tuned meter-reading vision models with a training workflow from ingestion to ONNX, deployed as scalable inference services on AWS.
Domains
Retail, Enterprise knowledge management, Industrial IoT, Business reporting
Education
BSc, Software Engineering (AI specialisation)