We are seeking an engineer who aspires to go beyond simple API calls and is passionate about deeply integrating AI capabilities with business scenarios. You will join our core R&D team, responsible for architecting AI training pipelines from scratch, building MCP (Model Context Protocol) services, and deploying AI Agents into real-world R&D workflows and business products.
Responsibilities
AI Skills Design & Training Pipeline Development
- Design and implement data pipelines and Skill encapsulations required for model training, transforming complex model capabilities into standardized interfaces for backend service invocation.
- Lead data cleaning, preprocessing, and feature engineering efforts to optimize model Fine-tuning for specific business domains.
- Iterate on model performance based on business feedback to ensure the accuracy and usability of AI outputs.
MCP (Model Context Protocol) Service Development
- Design and develop middleware services based on the Model Context Protocol (MCP) to manage contextual information during model interactions, ensuring state retention and logical routing in multi-turn conversations.
- Build reusable MCP components to guarantee low latency and high stability for AI applications under high concurrency.
AI Agent Implementation & Backend Integration
- Lead the development of AI Agents, designing agentic workflows (Planning/Tool Use) for specific scenarios such as automated code review, intelligent customer support, or R&D efficiency tools.
- Deeply integrate AI Agents into existing backend microservice architectures, bridging the gap between AI models and internal systems (e.g., databases, message queues, third-party APIs).
End-to-End AI Scenario Implementation
- Collaborate closely with product managers and backend engineers to identify pain points in R&D processes or business operations, proposing and implementing AI-driven solutions.
- Oversee the full delivery lifecycle from Proof of Concept (POC) to production, ensuring AI projects are not only feasible but also effectively utilized.
Qualifications
Essential Foundation: Solid Backend Experience
- Bachelor's degree or higher in Computer Science or a related field, with 3-5+ years of backend development experience.
- Proficiency in at least one mainstream backend language (Python/Go/Java), with a strong command of Linux environments and distributed system architectures.
- Deep practical experience with databases (SQL/NoSQL), caching (Redis), message queues, and other backend components.
Core Competencies: AI Engineering & Model Understanding
- AI Skills: Familiar with the full model training lifecycle; hands-on experience with data cleaning, model fine-tuning, or Skill encapsulation. A mere user of existing APIs will not suffice.
- MCP (Model Context Protocol): Understand the principles of context management in LLM interactions; experience developing services for session memory, state management, or dynamic prompt assembly.
- AI Agent: Familiar with frameworks like LangChain, AutoGen, or similar; understands principles such as the ReAct pattern and Chain of Thought (CoT); has practical experience developing AI Agents to solve specific problems.
Project Experience: Proven Track Record of Implementation
- Previous experience at an internet technology company with a proven track record of successfully applying AI (LLMs/Machine Learning) to enhance R&D efficiency (e.g., automated testing, intelligent debugging) or to power core business scenarios (e.g., recommendation systems, risk control, intelligent operations).
- A deep understanding of AI implementation, with the ability to derive technical solutions from business pain points rather than just stacking technologies.
Nice to Have
- High-quality open-source contributions related to AI on GitHub, or published technical blogs/papers.
- Knowledge of underlying LLM inference optimization or model compression techniques.
- Strong cross-team communication skills, capable of articulating the business value of AI solutions to non-technical stakeholders.
