AI Engineer (Machine Learning/LLM/API)

Job description

About the position

CLUEGA Company is looking for a AI Engineer with strong experience in Machine Learning, Deep Learning, NLP, LLM, RAG and AI Agent development, who can directly participate in researching, designing, building and deploying AI solutions into real products.

The right candidate is someone who has a solid technical foundation in AI/ML, understands how to work with real-world data, and is capable of moving from research, experiment, prototype to production deployment.

The candidate should be able to work with large-scale data from APIs, databases and internal systems, design suitable AI pipelines, and apply LLM-based solutions such as RAG, semantic search, AI Agent, function calling, tool calling, text classification, summarization, prediction and recommendation.

Responsibilities

• Design and build LLM-based agents with planning, tool calling, execution, observation, retry, and error handling.
• Develop tool connectors for internal systems, databases, files, dashboards, TikTok/Facebook/Google APIs, and third-party platforms.
• Build RAG pipelines for internal knowledge retrieval, including hybrid search, reranking, metadata filtering, and source-grounded responses.
• Implement memory, task state, checkpoints, and long-running workflow support for autonomous agents.
• Create guardrails for hallucination control, prompt injection defense, permission boundaries, and safe tool usage.
• Build evaluation frameworks for agent task success rate, retrieval quality, tool-call accuracy, latency, cost, and failure analysis.
• Collaborate with Product, Backend, QA, and Operations teams to convert real business workflows into AI agent flows.

Your skills and experience

Requirements

• 1-2+ years of experience as an AI Engineer, LLM Engineer, NLP Engineer, ML Engineer, or similar role.
• Strong Python skills and experience building production-grade AI/LLM applications.
• Hands-on experience with LLM tool calling, function calling, structured outputs, prompt engineering, and agentic workflows.
• Experience with LangChain, LangGraph, LlamaIndex, AutoGen, CrewAI, Semantic Kernel, or similar frameworks.
• Strong understanding of RAG, embeddings, vector search, hybrid search, reranking, and retrieval evaluation.
• Experience building backend APIs with FastAPI or similar frameworks.
• Familiarity with vector databases such as Qdrant, Milvus, Weaviate, FAISS, or similar.
• Understanding of common agent failure modes: hallucination, wrong tool use, stale data, prompt injection, context overflow, and unreliable execution.
• Strong ownership mindset and ability to work in a fast-moving startup environment.

Preferred Skills

• Experience building production AI agents, workflow automation systems, or Manus-like autonomous task systems.
• Experience with browser/web automation using Playwright, Selenium, or computer-use agents.
• Experience with MCP-like tool architectures, tool registries, permission systems, or multi-agent orchestration.
• Experience with TikTok, Facebook, Google APIs, ad tech, social media analytics, or content operation platforms.
• Experience with PostgreSQL, ClickHouse, Redis, Kafka, Docker, vLLM, TGI, Triton, or Ray Serve.
• Chinese language ability is a plus.

Tech Stack

Python, LLM APIs, open-source LLMs, LangChain/LangGraph/LlamaIndex, FastAPI, Vector DBs, PostgreSQL, ClickHouse, Redis, Docker, Playwright/Selenium, TikTok/Facebook/Google APIs.

Evaluation

Candidates may be asked to discuss or design:

• A Manus-like agent architecture for real business workflows.
• Tool schema design, permission control, retry logic, and error handling.
• RAG pipelines for internal documents, API knowledge, and frequently updated business data.
• Strategies for hallucination, stale data, prompt injection, context limits, and tool failure.
• Metrics for agent task success rate, retrieval quality, tool-call accuracy, latency, and cost.

Why you'll love working here

• Salary: 18M – 60,5M/month, depending on ability.
• Participate in social insurance, health insurance fully as prescribed.
• Working in a startup environment, the product develops quickly and has many opportunities to participate in AI system design from an early stage.
• Opportunity to build real AI products related to social media, advertising data, content data and platforms such as TikTok, Facebook, Zalo in the future.
• Opportunity to work from research to production, including RAG, AI Agent, LLM integration, model optimization and data-driven AI systems.
• Overtime with allowance