AI Agent / RAG Intern
招聘时间
即日起 至 2026-05-08
招聘单位
康宁
所在地区
上海
工作地点
上海市/上海市/浦东新区 康宁(上海)有限公司
学历要求
硕士
实习时长
每周4天，每天6小时
职责描述
Day to Day Responsibilities:
•Participate in developing AI Agent prototypes for internal scenarios; implement basic agent workflows (prompting, tool/function calling, memory/state) under guidance.
•Build and optimize RAG pipelines: data ingestion/cleaning, chunking strategy, embeddings, indexing, retrieval, re-ranking, and grounding/citation.
•Explore GraphRAG/knowledge-graph-assisted retrieval: help with schema design, entity/relation extraction, graph storage/query, and hybrid retrieval experiments.
•Assist in building evaluation datasets and metrics for RAG/Agent quality (e.g., Recall@k, MRR, answer correctness/groundedness), and iterate based on findings.
•Work with engineers to integrate components into services (e.g., simple APIs/scripts), write documentation, and present weekly progress updates.
•Continuously improve retrieval and generation quality through controlled experiments (prompting, query rewriting, retrieval parameters) and summarize learnings into best practices.
•Communicate effectively with cross-functional stakeholders; demonstrate independent thinking, proactive problem solving, and a positive, open personality.
Requirement：
•Currently pursuing a Master’s degree (enrolled) in Computer Science, Artificial Intelligence, Data Science, Software Engineering, or a related major.
•Familiar with AI Agent frameworks and RAG/GraphRAG concepts; candidates with hands-on project experience (course/research/industry/open-source) are preferred.
•Internship availability: 4+ days/week; 3+ months is preferred.
•Proficient in Python; able to write readable, testable code with basic engineering practices (logging, error handling, version control).
•Good understanding of RAG: document processing, chunking, embedding, indexing, retrieval & re-ranking, grounding/citation; end-to-end project experience is a plus.
•Familiar with mainstream AI Agent frameworks (e.g., LangChain / LlamaIndex / Semantic Kernel or similar) and agent design patterns (planning/orchestra
•Participate in developing AI Agent prototypes for internal scenarios; implement basic agent workflows (prompting, tool/function calling, memory/state) under guidance.
•Build and optimize RAG pipelines: data ingestion/cleaning, chunking strategy, embeddings, indexing, retrieval, re-ranking, and grounding/citation.
•Explore GraphRAG/knowledge-graph-assisted retrieval: help with schema design, entity/relation extraction, graph storage/query, and hybrid retrieval experiments.
•Assist in building evaluation datasets and metrics for RAG/Agent quality (e.g., Recall@k, MRR, answer correctness/groundedness), and iterate based on findings.
•Work with engineers to integrate components into services (e.g., simple APIs/scripts), write documentation, and present weekly progress updates.
•Continuously improve retrieval and generation quality through controlled experiments (prompting, query rewriting, retrieval parameters) and summarize learnings into best practices.
•Communicate effectively with cross-functional stakeholders; demonstrate independent thinking, proactive problem solving, and a positive, open personality.
Requirement：
•Currently pursuing a Master’s degree (enrolled) in Computer Science, Artificial Intelligence, Data Science, Software Engineering, or a related major.
•Familiar with AI Agent frameworks and RAG/GraphRAG concepts; candidates with hands-on project experience (course/research/industry/open-source) are preferred.
•Internship availability: 4+ days/week; 3+ months is preferred.
•Proficient in Python; able to write readable, testable code with basic engineering practices (logging, error handling, version control).
•Good understanding of RAG: document processing, chunking, embedding, indexing, retrieval & re-ranking, grounding/citation; end-to-end project experience is a plus.
•Familiar with mainstream AI Agent frameworks (e.g., LangChain / LlamaIndex / Semantic Kernel or similar) and agent design patterns (planning/orchestra
