I recently took a slight detour from traditional ML and started exploring Agentic AI and I’m learning through building projects in phases to really understand how things work from the barebones. Now, two projects deep I’m wondering how’s it is like in a professional setup.

In phase 1, I built a CLI research agent with a skeletal framework, and in phase 2, I added memory + RAG using chromaDB as my local knowledge base, tavily for web search, and DeepSeek as LLM.

[留言]

为什么值得关注

能改变理解方式，而不只是重复常识；符合当前抓取需求；它提供了新的理解或解释，而不只是表面观点

来源：reddit，领域：projects，保留分：0.60