Build an AI Research Agent
summary
Walks through building a working AI research agent that searches arXiv, returns ranked paper lists, generates concise summaries, and answers follow-up questions using a persisted knowledge base. Uses LangChain for orchestration, MongoDB Atlas (via Pongo and LangChain integration) as vector store and chat memory, Hugging Face datasets and embeddings, and Fireworks v1 as the LLM brain. Implements Chain of Thought and ReAct patterns, three tool endpoints (list, summarize, answer), Colab notebooks with runnable templates, and short-term memory persisted to MongoDB.
AI-written summary · about
chapters (11)
- 0:00Workshop introduction
- 2:32Understanding AI agents
- 6:05Agent use cases
- 8:53Reasoning and planning
- 12:43Agent memory components
- 14:36External tools for agents
- 15:36Building a research agent
- 18:05Setting up the environment
- 19:08Creating agent tools
- 22:46Developing the agent
- 25:20Adding memory and wrapping
related talks
02h 08mIntroduction to AI Agents
01h 23mBuild your own AI Agent
01h 12mBuild Your Own AI Chatbot with OpenAI & LangChain
53mHands-on with AI agents: From scratch to LangGraph
01h 27mBlazing the AI Trail: Using LangGraph to Conquer the Oregon Trail
40m4 pillars for building a modern AI application in JavaScript
34mJavaScript-Powered AI: Crafting Intelligent Systems with LLMs and Vector Databases
25mLangChain with Azure AI: Building LLM powered Applications