AI Scientist - LLM & AI Agent

We are looking for highly skilled NLP & LLM Experts to join our AI team and drive innovation in Agentic AI systems, focusing on reasoning, memory management, tool use, and multi-agent orchestration. This role is ideal for individuals who are passionate about building intelligent systems that go beyond traditional LLM applications — leveraging LangGraph, Model Context Protocol (MCP), Retrieval-Augmented Generation (RAG), and modular AI agents to power dynamic, context-aware applications.

You’ll help shape the next generation of intelligent automation across domains such as customer support, internal tooling, and knowledge orchestration.

Key Responsibilities

• Design and implement modular AI agents using frameworks such as LangGraph, enabling multi-turn reasoning, tool usage, and context retention.
• Build agentic workflows with Model Context Protocol (MCP) to manage dynamic memory, context injection, and action coordination.
• Develop and fine-tune LLMs (e.g., GPT-4, Qwen, LLaMA, Mistral) for downstream tasks such as multi-agent collaboration, structured generation, and reasoning.
• Integrate Retrieval-Augmented Generation (RAG) architectures with vector databases (e.g., FAISS, Milvus) to support context-aware responses and tool routing.
• Engineer prompts and multi-step reasoning chains for LLMs in agent-based environments.
• Deploy and optimize open-source LLMs using vLLM, Triton, or Ollama, ensuring low-latency inference at scale.
• Translate core NLP capabilities into production-ready agent behaviors, collaborating with engineering and product teams.
• Stay at the forefront of agentic AI research, contributing to internal frameworks and incorporating cutting-edge ideas from open source and academia.
• Present project outcomes and model behaviors to both technical and non-technical stakeholders to guide product strategy.

Basic Qualifications

• Master’s, or PhD in Computer Science, Artificial Intelligence, or a related field.
• 2+ years of experience in AI or NLP-focused roles, including production deployment of LLM-driven systems.
• Proficiency in Python and libraries such as Transformers, LangGraph, LangChain, and PyTorch.
• Strong knowledge of prompt engineering, context management, and multi-agent orchestration.
• Hands-on experience deploying and scaling LLMs in production environments.
• Familiarity with retrieval mechanisms, vector search, and memory architectures in agent workflows.
• Able to use Chines and English as daily working lanague to work with chinese speaking skatesholder