Name: THE AI EVALUATION PLAYBOOK: How to Choose the Right Metrics, Methods and Evaluators for Modern AI Systems (en Inglés)
SKU: 69158870
Price: 82139.00 COP
Availability: InStock
Author: Thomas, Nicholas
ISBN: 9798160091693

portada THE AI EVALUATION PLAYBOOK: How to Choose the Right Metrics, Methods and Evaluators for Modern AI Systems (en Inglés)

THE AI EVALUATION PLAYBOOK: How to Choose the Right Metrics, Methods and Evaluators for Modern AI Systems (en Inglés)

Formato

Libro Físico

Editorial

Independently published

Autor

Thomas, Nicholas

Idioma

Inglés

N° páginas

278

Encuadernación

Tapa Blanda

ISBN13

9798160091693

THE AI EVALUATION PLAYBOOK: How to Choose the Right Metrics, Methods and Evaluators for Modern AI Systems (en Inglés)

Thomas, Nicholas (Autor) · Independently published · Tapa Blanda

Libro Nuevo Importado

Envío: 16 a 22 días háb.

$ 164.277$ 82.139

-50%

Costos de importación incluídos en el precio ✅

Libro Nuevo

Quedan 100 unidades

$ 82.139

Llega entre el 11 Dic y el 22 Dic a Bogota, Cundinamarca. Seleccionar ubicación

Agregar a lista de deseos

Reseña del libro "THE AI EVALUATION PLAYBOOK: How to Choose the Right Metrics, Methods and Evaluators for Modern AI Systems (en Inglés)"

An AI system can pass its tests—and still be unsafe to trust. An answer may sound right but lack evidence. A RAG system may find the right source yet still mislead. An agent may finish a task while crossing a permission boundary. Even the evaluator may be wrong. The real question is not “Did it score well?” It is “Do we have enough evidence to trust it?” The AI Evaluation Playbook gives you a practical way to answer that question. Learn to: Define intended behavior and acceptance criteria before choosing metrics. Match evaluation methods and evaluators to the system, task, and risk. Test LLM judges and other evaluators for validity, bias, and reliability. Evaluate RAG, memory, tool use, agent autonomy, multi-agent coordination, and multimodal AI. Turn findings into release gates, production monitoring, and continuous assurance. An end-to-end enterprise case takes you from sample data and a rubric to release decisions and monitoring. For AI leaders and practitioners deciding what AI is ready to do—and what evidence is strong enough to trust.