Future AGI Cookbooks: Guides and Tutorials

Runnable recipes for evaluation, tracing, simulation, and optimization, organized by what you are building and by platform feature.

Start with a quickstart, or jump straight to the recipes for what you’re building or the platform feature you need.

Start Here

Score LLM outputs for hallucination, toxicity, and custom criteria

Create, edit, and evaluate a dataset from the dashboard

Upload a CSV, run batch evals, and download scored results

Run built-in eval templates with the ai-evaluation package

Deploy the full open-source stack locally in five minutes

Track latency and cost trends, then set threshold alerts

Evaluate and simulate conversational agents, 7 recipes

Score retrieval, generation, and grounding, 9 recipes

Test voice agents with scripted call simulations, 2 recipes

Instrument and evaluate multi-agent and tool-calling systems, 5 recipes

Score text, image, audio, and PDF outputs, 5 recipes

Build and evaluate text-to-SQL agents, 2 recipes

Instrument, connect, and debug traces, 8 recipes

Write custom metrics and run evals at scale, 5 recipes

Generate, import, and annotate datasets, 5 recipes

Version, compare, and optimize prompts, 10 recipes

Was this page helpful?

Questions & Discussion