The lab notebook
Field notes on AI engineering
824 articles · page 24 of 28
LangFlow Self-Hosting: Docker Setup and Configuration
A practical guide to self-hosting LangFlow with Docker Compose, covering Postgres persistence, environment variables, auth, and production hardening.
LangFlow Team Collaboration: Sharing Flows Across a Team
A practical guide to sharing LangFlow flows across a team without losing track of versions, credentials, or who changed what.
LangFlow Templates: Starting from Pre-Built Flows
Learn how LangFlow templates let you skip boilerplate and start building AI apps from working, pre-built flows in minutes.
LangFlow Version Control: Tracking Changes to Visual Flows
Learn how to git-diff LangFlow's JSON exports, structure commits, and recover from breaking changes to visual AI flows.
LangFlow vs Flowise: Comparing Visual LangChain Builders
A practical, no-hype comparison of LangFlow and Flowise for building LLM workflows visually, covering architecture, components, deployment, and when to pick each.
Building a LangGraph Agent That Writes and Runs Its Own Tests
Learn how to build a LangGraph agent that generates unit tests, executes them in a sandbox, and loops on failures until the code passes.
LangGraph for Code Generation Agents: Plan, Write, Test, Fix
Learn how to build a self-correcting code generation agent in LangGraph with plan, write, test, and fix nodes.
LangGraph Common Pitfalls: Infinite Loops and State Bloat
LangGraph agents fail in predictable ways — here's how to spot and fix infinite loops, runaway state, and silent recursion limits before production does it for you.
LangGraph Conditional Edges: Building Branching Agent Logic
Learn how LangGraph conditional edges let agents route dynamically between nodes, with working add_conditional_edges code and common pitfalls.
LangGraph for Content Moderation Pipelines
Learn how to design a stateful, auditable content moderation pipeline in LangGraph that combines classifiers, LLM judgment, and human review.
LangGraph Cost Tracking Across a Multi-Node Graph
Learn how to track LLM token costs per node across a LangGraph graph using callbacks, state, and checkpoints.
LangGraph for Customer Support Escalation Flows
Learn how to build production-grade customer support escalation flows with LangGraph: state schemas, conditional routing, human handoff, and checkpointing.
LangGraph Deployment: Serving a Graph as a Production API
Learn how to take a LangGraph agent from a notebook prototype to a durable, streaming production API using LangGraph Platform and FastAPI.
LangGraph Error Recovery: Handling Node Failures Gracefully
A practical guide to catching node failures, retrying safely, and routing around broken steps in LangGraph without losing state.
LangGraph for Financial Analysis Agents
Learn how to build LangGraph financial analysis agents with stateful graphs, tool calling, reflection loops, and human-in-the-loop review for finance work.
LangGraph Interrupt Patterns Beyond Human Approval
Master langgraph interrupt patterns beyond approve/reject: tool editing, input collection, validation loops, webhook waits, and re-execution gotchas.
LangGraph for Legal Document Review Agents
Build a LangGraph legal review agent with typed state, conditional routing, human-in-the-loop approval, and checkpointed audit trails for contracts.
LangGraph for Long-Running Workflows: Pausing and Resuming Days Later
Build LangGraph long running workflows that pause for days and resume exactly where they stopped, using checkpointers, interrupts, and Postgres persistence.
LangGraph Migration Guide: Moving from Plain LangChain Agents
A practical, code-first guide to migrating plain LangChain agents to LangGraph, covering state, control flow, memory, and rollout strategy.
LangGraph for Multi-Agent Supervisor Patterns
A practical guide to building supervisor-led multi-agent systems in LangGraph, with routing logic, state design, and failure handling.
LangGraph for Onboarding Assistants: Multi-Step Guided Flows
Learn how LangGraph's stateful graphs make onboarding assistants reliable, resumable, and easy to debug compared to plain prompt chains.
LangGraph Persistence Backends Compared: SQLite, Postgres, Redis
Compare LangGraph persistence backends — SQLite, Postgres, and Redis checkpointers — with real code, trade-offs, and a decision framework for production.
LangGraph Reducers: Controlling How State Updates Merge
Learn how LangGraph reducers control state merging with Annotated types, operator.add, add_messages, and custom merge functions for parallel agents.
LangGraph for Research Agents: Iterative Search and Synthesis
Build LangGraph research agents that search, reflect, and re-search in a loop, then synthesize cited answers — with full Python code and stop conditions.
LangGraph for Sales Qualification Agents
Learn how to build a stateful, multi-step sales qualification agent with LangGraph that scores leads, asks smart follow-ups, and routes hot prospects to humans.
LangGraph Scaling: Running Many Graph Instances Concurrently
Learn how LangGraph scaling really works: run hundreds of concurrent graph executions with async invocation, Postgres checkpointers, and backpressure.
LangGraph Send API: Dynamic Parallel Node Execution
Learn how LangGraph's Send API dynamically fans out parallel node executions at runtime, with real code for map-reduce, subgraphs, and state merging.
LangGraph State Schemas: Designing Your Agent's Shared Memory
A practical guide to designing LangGraph state schemas with TypedDict, reducers, and multi-agent patterns that actually hold up in production.
LangGraph Streaming: Returning Intermediate Steps to Users
Learn how to stream intermediate steps from a LangGraph agent using .stream() and .astream_events() so users see reasoning, tool calls, and tokens live.
LangGraph Studio: Visual Debugging for Agent Graphs
Learn how LangGraph Studio turns invisible agent state into a visual graph you can inspect, replay, and fix node by node.