{ "item_title" : "Loop Engineering", "item_author" : [" Fabio Correa "], "item_description" : "Most books about working with AI review the model. This one reviews the work around the model, and finds it wanting in specific, fixable ways.A clinician reviews a scan. An analyst tests a forecast. A writer revises a draft. A founder adjusts a plan. Different materials, one recurring cycle: intent, action, feedback, memory, correction, return. Loop Engineering treats that cycle as a design problem, and asks what a well-built loop owes the person who remains accountable for the result.Its opening failure list will be familiar to anyone who has shipped AI-assisted work: no clear definition of done, no check independent of the work itself, no memory of what was already tried, no record of the shortcuts taken along the way. The book traces how practitioners in four fields arrived at the same remedies independently, then draws the recurring mechanisms into one architecture.The seven design commitmentsVerifiable stop: every iteration ends with a check that passes or failsMaker/checker: the work is verified by something other than its makerExternalized memory: state lives outside the conversation, where it survivesMateriality gating: the stakes set how much loop the work earnsFrame before solve: restate the problem before solving itVisible verification: the check is part of the record, not a claim about itDebt accumulation: failed branches, shortcuts, and open questions are logged, not lostHow a working reader would use itFrame before solve becomes the first five minutes of a brief: the problem restated in writing before anyone prompts a model.Maker/checker becomes a standing rule for deliverables: a forecast, a dashboard, or a draft is verified by a test, a second pass, or a second person, never only by the pass that produced it.Externalized memory becomes the project record that lives outside the chat window, so the third session starts where the second one ended.The materiality dial becomes a triage habit: a routine summary gets a light loop; a board-facing number gets an evidence gate and a named approver.Debt accumulation becomes the log nobody wants to keep and everybody needs: the shortcuts, the failed branches, the open questions, written down while they are still cheap.The second edition adds the strategic harness (Chapter 17): the same discipline at the scale of a team or a department, with a one-page Harness Canvas for the controls, evidence, and approvals an organization puts its name on.Who should read itLeaders, analysts, builders, educators, writers, and researchers who direct AI-assisted work and remain accountable for the result. It is written for readers who want a discipline that survives the next tool change.Delegate the work. Never the accountability.Loop Engineering is the foundation volume of the Applied AI series. Fabio Correa directs business analytics at Microsoft and teaches AI-infused business analytics at Virginia Tech. His companion book, The Defensible Decision, applies the same discipline to executive decision-making.", "item_img_path" : "https://covers2.booksamillion.com/covers/bam/9/79/818/460/9798184606361_b.jpg", "price_data" : { "retail_price" : "29.99", "online_price" : "29.99", "our_price" : "29.99", "club_price" : "29.99", "savings_pct" : "0", "savings_amt" : "0.00", "club_savings_pct" : "0", "club_savings_amt" : "0.00", "discount_pct" : "10", "store_price" : "" } }

Overview

Customers Also Bought

Details