Building LLM Agents (Hackathon)
summary
Shows how to build a custom LLM agent using LangChain, demonstrating concrete implementations inspired by AutoGPT, BabyAGI and similar simple agent patterns. Covers handing decision-making and task execution to an LLM and explains underlying concepts such as ReAct, Plan & Execute, and CAMEL while assembling an agent that can converse and autonomously handle complex tasks.
AI-written summary · about
chapters (18)
- 0:00LLM agents overview
- 5:08What are LLM agents
- 9:18Why agents matter now
- 18:54Prompt engineering for agents
- 22:46LangChain for agents
- 27:45Live demo: simple agents
- 35:34Agent types and REACT
- 40:09Agent tools: search, calculator
- 49:24Agent safety and interaction
- 54:44Long-term memory with vector DBs
- 1:00:59Document chatbots with LangChain
- 1:10:20Agents for SQL and structured data
- 1:18:05Plan-and-execute agents
- 1:25:23BabyAG and autonomous agents
- 1:30:19Success factors for domain agents
- 1:32:25Teamwork with Camel agents
- 1:38:54Agents for creative tasks
- 1:40:39Future of agents
related talks
50mIntelligent Agents, the New Frontier of LLMs
41mBuilding LLM Agents – Challenges and Learnings
36mLLMs and AI agents demystified
35mBuilding LLM-based Agents
01h 59mLLM applications with LangChain
01h 34mA Practical Starter's Guide to building LLM based projects
02h 08mIntroduction to AI Agents
09mBuilding LLM Agents with Tool Use