This article explores how to create AI agents for web search, financial analysis, reasoning, and retrieval-augmented generation using phidata and the Ollama local LLM.
The article covers the installation process of UV and the creation of a project folder.
The web search agent is created using DuckDuckGo search engine.
The finance agent is created using YFinance tools for stock price, analyst recommendations, company info, and company news.
The agent team integrates the agents created in the previous steps for web search and finance.
The inference agent is designed to perform reasoning using a given task.
The RAG agent is created to use a PDF knowledge base and a LanceDB vector db.