didilili/ai-agents-from-zero

Combines runnable code examples, low-code platforms (Coze/Dify), and open-source frameworks (LangChain/LangGraph) across a unified learning path from prompt engineering through enterprise RAG/Agent systems and fine-tuning. Emphasizes practical deployment patterns—vector + sparse + knowledge graph retrieval, streaming with memory management, human handoff workflows—grounded in real e-commerce scenarios (merchant copilot, customer support triage, NL-to-SQL analytics). Includes interview question bank organized by job competencies, making it directly applicable to both freelance project delivery and formal hiring calibration.