menu
{ "item_title" : "Build AI Agents for Software QA", "item_author" : [" Anup Naik "], "item_description" : "Stop copying AI answers into your test plan. Build agents that show their evidence.Language models can draft test cases, summarise logs and write reports in seconds. They can also invent a quote, soften bad news or follow instructions hidden in a bug ticket. This book shows QA engineers, test leads and SDETs how to use that speed without giving up the discipline that makes testing trustworthy.You build 21 working agents in Python: a first warm-up agent, then 20 hands-on agents. Every one has a one-page spec card, a clear output format, code that checks the model's claims against the source, and a named human checkpoint.What you will buildTest design: requirement reviewers, test case generators, test data and traceability agentsExecution and automation: exploratory session companion, impact analysis, API contract drift, payload checks, accessibility and visual driftDefects and triage: bug report writer, log and root-cause summariser, regression mapping, flaky test triageBeyond function: performance scenarios, security test ideas, evaluating LLM outputDecisions: release readiness scorecard and audience-specific test reportsThen you run them as a team. The final part covers orchestrating agents with limits and approvals, attacking your own agents with a compromised model to see what still stops it, measuring value with honest uncertainty, and rolling out in stages. A capstone shows how to generate your own next agent.Companion code included. The book comes with a tested code pack (256 automated tests) that runs offline with a stand-in model, so you can learn without an account. Connect a model provider when you are ready. All sample data is invented.You will need: basic Python and a terminal. No machine-learning background required.The book does not promise results. It teaches you to measure them on your own team.", "item_img_path" : "https://covers1.booksamillion.com/covers/bam/9/79/816/567/9798165676956_b.jpg", "price_data" : { "retail_price" : "34.99", "online_price" : "34.99", "our_price" : "34.99", "club_price" : "34.99", "savings_pct" : "0", "savings_amt" : "0.00", "club_savings_pct" : "0", "club_savings_amt" : "0.00", "discount_pct" : "10", "store_price" : "" } }
Build AI Agents for Software QA|Anup Naik

Build AI Agents for Software QA : Hands-On Agents for Test Design, Automation, Triage and Release Decisions

local_shippingShip to Me
In Stock.
FREE Shipping for Club Members help

Overview

Stop copying AI answers into your test plan. Build agents that show their evidence.

Language models can draft test cases, summarise logs and write reports in seconds. They can also invent a quote, soften bad news or follow instructions hidden in a bug ticket. This book shows QA engineers, test leads and SDETs how to use that speed without giving up the discipline that makes testing trustworthy.

You build 21 working agents in Python: a first warm-up agent, then 20 hands-on agents. Every one has a one-page spec card, a clear output format, code that checks the model's claims against the source, and a named human checkpoint.

What you will build

  • Test design: requirement reviewers, test case generators, test data and traceability agents
  • Execution and automation: exploratory session companion, impact analysis, API contract drift, payload checks, accessibility and visual drift
  • Defects and triage: bug report writer, log and root-cause summariser, regression mapping, flaky test triage
  • Beyond function: performance scenarios, security test ideas, evaluating LLM output
  • Decisions: release readiness scorecard and audience-specific test reports

Then you run them as a team. The final part covers orchestrating agents with limits and approvals, attacking your own agents with a compromised model to see what still stops it, measuring value with honest uncertainty, and rolling out in stages. A capstone shows how to generate your own next agent.

Companion code included. The book comes with a tested code pack (256 automated tests) that runs offline with a stand-in model, so you can learn without an account. Connect a model provider when you are ready. All sample data is invented.

You will need: basic Python and a terminal. No machine-learning background required.

The book does not promise results. It teaches you to measure them on your own team.

This item is Non-Returnable

Details

  • ISBN-13: 9798165676956
  • ISBN-10: 9798165676956
  • Publisher: Independently Published
  • Publish Date: October 2026
  • Dimensions: 9 x 6 x 0.92 inches
  • Shipping Weight: 1.33 pounds
  • Page Count: 454

Related Categories

You May Also Like...

    1

BAM Customer Reviews