LLM/RAG Infra Engineer — Multi-Agent AI

SleekFlow; Hong Kong; Hybrid; HKD 420,000 - 640,000; Full time

Benefits offered by this job

Stock options

13th-month salary

Medical insurance

Transport allowance

Education allowance

Birthday leave

Flexible work-from-home

Job summary

SleekFlow, an AI suite provider, seeks a skilled LLM/RAG/Multi-Agent Infrastructure Engineer in Hong Kong. You will design, deploy, and optimize systems that integrate LLMs via APIs, build efficient RAG pipelines, and enable multi-agent coordination across scalable AI applications.

The role requires 1–2 years of experience with RAG/LLM development, strong knowledge of vector databases and embedding techniques, and a proactive approach to collaboration with AI researchers and software engineers.

Qualifications

• Hands-on experience in integrating LLM APIs (e.g., OpenAI, Anthropic, Cohere) and building RAG pipelines with vector search.
• Familiarity with multi-agent systems and their infrastructure requirements.
• Experience in designing, indexing, and maintaining structured and unstructured knowledge bases, including ontology design and embedding techniques.
• Experience with ML model versioning, deployment pipelines, and monitoring tools.
• Proficiency in designing clear, structured prompts and leveraging few-shot learning and chain-of-thought reasoning.

Responsibilities

• LLM Evaluation and model selection for specific applications and use cases.
• Develop and scale RAG architectures, integrating vector databases and embedding models.
• Design robust multi-agent coordination infrastructure for distributed decision-making.
• Build knowledge bases and automate document indexing for high-quality outputs.
• Implement observability, monitoring, and security controls for AI workloads.

Skills

LLM APIs

RAG pipelines

Multi-Agent infra

Knowledge base engineering

CI/CD & MLOps

Prompt engineering

Tools

FAISS

Weaviate

Pinecone

Job description

SleekFlow, an AI suite provider, seeks a skilled LLM/RAG/Multi-Agent Infrastructure Engineer in Hong Kong. You will design, deploy, and optimize systems that integrate LLMs via APIs, build efficient RAG pipelines, and enable multi-agent coordination across scalable AI applications.

The role requires 1–2 years of experience with RAG/LLM development, strong knowledge of vector databases and embedding techniques, and a proactive approach to collaboration with AI researchers and software engineers.