Generative AI & LLM Engineering

소개

This skill equips Claude with the technical expertise to design and implement complex generative AI systems, focusing on production-grade Retrieval-Augmented Generation (RAG), agentic workflows, and sophisticated prompt engineering. It provides deep implementation patterns for frameworks like LangChain and LlamaIndex, integration guidance for vector databases such as Chroma and Pinecone, and strategies for implementing robust monitoring and evaluation to ensure AI application reliability and performance in real-world environments.

주요 기능

• Sophisticated prompt engineering patterns including Few-Shot and Chain of Thought
• 1 GitHub stars
• Evaluation and observability setup using LangSmith and custom metrics
• Structured output parsing for reliable data extraction and analysis
• Advanced RAG implementation with hybrid search and contextual compression
• Development of autonomous agents with structured tool use and Pydantic validation

사용 사례

• Building a specialized knowledge base assistant using private document repositories
• Developing autonomous data analysis agents that can execute SQL and perform calculations
• Implementing automated evaluation pipelines to benchmark LLM response quality