Patterns for Building AI Agents - Tapa blanda

Sinopsis

The gap between prototype and production is where most AI teams stumble. Written by the co-creator of a popular agent framework, Patterns for Building AI Agents captures practical strategies emerging in the year of agents:

• Agent design patterns: Evolving architectures, creating dynamic agents, and building human-in-the-loop workflows
• Context engineering: Parallelization, context compression, and avoiding failure modes
• Eval workflows: Creating eval suites, cross-referencing failure modes with metrics, and leveraging domain experts
• Security fundamentals: Preventing prompt injection, sandboxing code execution, and implementing agent access control

"Sinopsis" puede pertenecer a otra edición de este libro.