Talki Academy
🔗

LangChain and LangGraph in Production

Intensive technical training for developers who want to master LangChain and LangGraph, the reference open-source frameworks for building AI agents and multi-step workflows in production. From LCEL to multi-agent systems, learn to deploy reliable, scalable, and observable applications.

Duration
2 days
Level
Intermediate
Price
9.99 EUR/month (all courses included)
Max group
12 participants

What you will learn

+Master LCEL (LangChain Expression Language) to compose robust chains
+Build production agents with persistent memory and custom tools
+Orchestrate complex workflows with LangGraph (state, routing, cycles)
+Deploy with FastAPI, streaming, caching, and monitoring
+Manage LLM costs, optimize latency, and implement resilience
+Observe and debug with LangSmith in production

Course program

Module 1: LCEL: LangChain Expression Language

3h00
  • LCEL vs raw API calls: when to use abstraction?
  • Composition with the pipe operator: prompt | llm | parser
  • Patterns: sequential, parallel, conditional chains
  • Retry, fallback, timeout for production
  • Workshop: resume analysis pipeline (summary + skills + level)

Module 2: Production agents: memory and tools

3h30
  • Conversational memory with Redis (multi-user persistence)
  • Create custom business tools (Tool, StructuredTool, Pydantic)
  • Error handling, retry logic, circuit breakers for external APIs
  • Observability with LangSmith: traces, debugging, evaluation
  • Workshop: customer support agent with knowledge base

Module 3: LangGraph: complex workflows with state

4h00
  • LangGraph vs ReAct agents: explicit control vs LLM-driven
  • State management: TypedDict, Annotated, operator.add
  • Conditional routing: conditional_edges, quality-based retry
  • Human-in-the-loop with checkpoints (MemorySaver, RedisSaver)
  • Workshop: company data enrichment pipeline

Module 4: Deployment and production patterns

3h30
  • FastAPI with streaming (Server-Sent Events)
  • Cost management: Redis caching, semantic cache, prompt optimization
  • Monitoring: Prometheus, Grafana, metrics (latency, cost, errors)
  • Circuit breaker for external dependencies
  • Production checklist: health checks, rate limiting, secrets, logs

Ready to get started?

9.99 EUR/month — All courses included, cancel anytime

Request a quoteView all courses