Skip to content

Agentic AI Learning Path

Sequential roadmap from fundamentals to production

2026 Edition • ~200-400 hours total • Updated February 2026

Progress0 completed / 14 total

How to Use This Roadmap

  • Click steps to mark as completed and track your progress
  • Follow the order — each step builds on previous knowledge
  • Checkpoints are milestone projects to validate learning
  • Estimated times assume 10-15 hrs/week of focused study
  • Skip steps you already know — they count toward progress but are tracked separately
  • Click "Show details" to see learning objectives and resources

Phase 1: Foundations

4-6 weeks
Prerequisites: Basic programming experience helpful but not required • Familiarity with command line
0 completed / 3
1

Python & API Basics

Beginner
2 weeks

Core programming skills for AI development

Python 3.10+requests/httpxasync/awaitJSON handlingEnvironment management
View lesson →
2

Prompt Engineering

Beginner
1-2 weeks

Master the art of communicating with LLMs

Chain-of-ThoughtFew-shot promptingRole promptingOutput formattingPrompt iteration
View lesson →
3

LLM APIs & Function Calling

★ CHECKPOINTIntermediate
1-2 weeks

Connect to and orchestrate LLM providers

OpenAI APIAnthropic APIFunction/tool callingStreamingError handling
View lesson →

Phase 2: Core Agent Skills

6-8 weeks
Prerequisites: Completed Phase 1 • Comfortable with async Python • Basic understanding of LLM APIs
0 completed / 3
4

Tool Use & Integration

Intermediate
2 weeks

Extend agents with external capabilities

Tool schemasAPI integrationFile I/OCode executionError recovery
View lesson →
5

Agent Architectures

Intermediate
2 weeks

Understand core patterns for autonomous agents

ReAct patternPlan-and-ExecuteSelf-reflection loopsDecision policiesGoal decomposition
View lesson →
6

Memory Systems

★ CHECKPOINTIntermediate
2 weeks

Give agents persistent context and recall

Conversation memoryVector embeddingsSemantic searchMemory typesContext window management
View lesson →

Phase 3: Production Patterns

6-8 weeks
Prerequisites: Completed Phase 2 • Experience building basic agents • Understanding of vector embeddings
0 completed / 3
7

RAG Systems

Intermediate
2-3 weeks

Ground agents in your own data

Document chunkingEmbedding modelsVector storesHybrid searchRe-rankingQuery refinement
View lesson →
8

Agent Frameworks

Intermediate
2 weeks

Leverage production-ready tooling

LangChainLangGraphCrewAIState machinesGraph-based workflows
View lesson →
9

Orchestration & Workflows

★ CHECKPOINTAdvanced
2 weeks

Build reliable, observable pipelines

DAG designError handlingRetriesConditional logicState managementEvent triggers
View lesson →

Phase 4: Hardening

4-6 weeks
Prerequisites: Completed Phase 3 • Working agent in development • Understanding of production systems
0 completed / 3
10

Security & Governance

Advanced
2 weeks

Protect agents and users from harm

Prompt injection defenseInput validationOutput filteringRBACSandboxingAudit logging
View lesson →
11

Monitoring & Evaluation

Advanced
1-2 weeks

Observe, measure, and improve agent performance

LangSmith/LangfuseTracingMetricsA/B testingHuman feedback loopsBenchmarking
View lesson →
12

Deployment & Scaling

★ CHECKPOINTAdvanced
1-2 weeks

Ship agents to production

DockerFastAPIKubernetes basicsRate limitingCachingCost optimization
View lesson →

Phase 5: Advanced

Ongoing
Prerequisites: Completed Phases 1-4 • Production agent experience • Solid foundation in all core skills
0 completed / 2

Multi-Modal & Computer Use

ADVANCEDAdvanced
Ongoing

Cutting-edge agent capabilities

Vision modelsDocument understandingBrowser automationComputer useAudio/video processing
View lesson →

Multi-Agent Systems

ADVANCEDAdvanced
Ongoing

Coordinate teams of specialized agents

Agent communicationTask delegationConsensusMCP protocolA2A protocol
View lesson →

Building an AI agent but stuck on deployment?

I can help →