Use Case Cookbooks
Cookbooks organized by what you are building: chat and support agents, RAG, voice agents, multi-agent systems, content quality, and text-to-SQL.
Pick the section that matches what you are building; each recipe is a complete, runnable walkthrough.
Chat & Support Agents
Multi-Turn Conversation Eval
Score complete customer conversations, not single turns
Session-Based Observability
Group traces by session and user across turns
LangGraph Agent Observability
Trace a support agent and find which layer fails
Chat Simulation with Personas
Run multi-persona conversations against your agent via SDK
Chat Simulation & Fix My Agent
Simulate chats, then diagnose failures with Fix My Agent
Prompt Workbench Simulation
Simulate from the dashboard without SDK or code
End-to-End Agent Testing
Test and fix a chat agent with simulated conversations
RAG & Document Q&A
RAG Evaluation
Score retrieval and generation separately to localize failures
Hallucination Detection
Catch ungrounded answers with faithfulness and groundedness evals
Evaluating RAG Applications
End-to-end quality checks for a RAG pipeline
RAG Chatbot Trustworthiness
Measure whether answers stay inside retrieved context
Decrease RAG Hallucinations
Tune chunking and retrieval until hallucinations drop
LangChain RAG Evaluation
Evaluate a LangChain RAG stack with Future AGI
LlamaIndex PDF RAG Chatbot
Build and observe a PDF chatbot on LlamaIndex
MongoDB Atlas RAG Chatbot
PDF RAG on MongoDB Atlas vector search
Knowledge Base
Upload documents and query them with the SDK
Voice Agents
Voice Simulation
Define agents and personas, then run scripted call tests
Voice Agent Simulate SDK
Test a voice agent with the Agent Simulate SDK
Multi-Agent & Tool Use
CrewAI Research Team
Observe a multi-agent CrewAI research system
Google ADK Error Feed
Surface failures from Google ADK multi-agent traces
Tool-Calling Agent Simulation
Simulate and trace an agent that calls tools
Agent Function Calling Eval
Score function-call choices and response quality
LangChain & LangGraph Observability
Instrument LangChain and LangGraph applications
Content & Multimodal
Meeting Summarization Eval
Score summaries for coverage and faithfulness
AI SDR Outreach Eval
Evaluate outreach messages for quality and tone
Image Evaluation
Score AI-generated images with the SDK
Multimodal Evaluation
Evaluate images, audio, and PDF inputs
Tone, Toxicity & Bias Evals
Detect tone, toxicity, and bias in outputs
Text-to-SQL
Questions & Discussion