Chapter 2. The AI Product Development Lifecycle

Unlike traditional products, AI-driven products introduce a unique blend of code, data, algorithms, and user experience, creating a distinctive development narrative. The AI Product Development Lifecycle (AIPDL)1 captures the phases of developing an AI-powered product while ensuring the product meets users’ needs and will find market fit. In this chapter, I will introduce two types of AI products, 0-to-1 and 1-to-n, and detail how these products are developed following the phases of the AIPDL. Most AI products will follow the AIPDL quite closely, but depending on the product type, the time needed in each phase may vary.

Types of AI Products

The AIPDL depends on the type of AI product you are making. If you work at an early-stage startup, you may work on 0-to-1 products, applying an emerging technology or model to a new product to create an experience that did not exist before. If you work at a more established organization, you may also work with ...

Get Building AI-Powered Products now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.