Phase 6 of 9

Frameworks & Agents

Days 51-60 (10 days)

Standing on the Shoulders of Giants

You've built everything from scratch. Now learn when to use frameworks like LangChain and LlamaIndex - and when not to. Then dive into the frontier: AI agents that decide what to do autonomously.

By the end of this phase, you'll build a research assistant that searches the web, reads sources, and writes reports with citations.

What Problems Will You Solve?

Boilerplate
500 lines of RAG code - every project starts from scratch
Complexity
Sometimes frameworks add more complexity than they solve
Agents
Chains are hardcoded but tasks need dynamic decision-making
Multi-Agent
One agent can't do everything - need specialized teams
Live Data
LLMs have knowledge cutoffs - need web search tools

Daily Schedule

Day 51

LangChain Fundamentals

LCEL, PromptTemplates, chains, output parsers
Day 52

LangChain RAG

Document loaders, text splitters, retrievers
Day 53

LlamaIndex Fundamentals

LlamaHub, indices, query engines
Day 54

Agents: Concept & Tool Use

ReAct pattern, tool binding, decision loops
Day 55

LangGraph: Stateful Agents

State management, cycles, human-in-the-loop
Day 56

Multi-Agent Systems

CrewAI, roles, delegation, collaboration
Day 57

Web Search & Live Data Tools

Tavily, SerpAPI, real-time information
Day 58

Code Interpreter Patterns

Sandboxed execution, data analysis agents
Day 59

Framework Best Practices

When to use frameworks, over-engineering dangers
Day 60

Project: Research Assistant

Autonomous agent that researches and writes reports
← Phase 5 Start Day 51 →