Get Things Done with AI: A Practitioner's Guide to AI-Native Knowledge Work - Ravish Agrawal

Get Things Done with AI is the practical guide for working professionals who have tried AI, found it promising, and now want to use it seriously. Written for analysts, managers, consultants, operators, and individual contributors across every function, it occupies the productive middle ground between abstract speculation and shallow tool lists: operational enough to act on tomorrow morning, yet structural enough to remain relevant as the technology continues to evolve.

The book unfolds in four parts that build on one another. Part One establishes a mental model for AI-native problem solving, including the five-step Build Loop and the discipline of asking sharp questions. Part Two maps the modern AI stack (models, skills, the Model Context Protocol, and agents), equipping readers with the architecture to navigate any workflow. Part Three applies these foundations to the three core activities of knowledge work: investigating data, producing multi-voice communications, and building custom tools when off the shelf options fall short. Part Four addresses the discipline required to sustain AI native work over time, covering how to extract genuine disagreement from AI and how to graduate ad hoc experiments into reliable infrastructure.

Rounding out the book are profession specific playbooks for six common functions: finance, sales, marketing, product, customer success, and human resources. Readers can work through the book from cover to cover or jump directly to the playbook that matches their role and use the chapters as a reference from there. Either way, Get Things Done with AI delivers a durable framework for the biggest productivity shift most knowledge workers will experience in their careers.