Back to homepage The PM's handbook for building AI features 🪨

Before we dive in, let’s get clear on the 4 keys to success that should be at the forefront of your mind:

1️⃣ AI product building disrupts the traditional product development process

Building with AI means incorporating a non-deterministic technology from a third-party provider into your product. This brings a new paradigm and differs significantly from traditional product building. It's natural for teams and organizations to face challenges during the learning curve—though it's often not as steep as one might expect.

The key is to adopt the right mindset: be flexible, start building, and keep moving forward.

2️⃣ Normalize iterations, iterations, and more iterations

Building with AI is a highly iterative process that requires a lot of trial and error. The value of your AI product hinges on the quality of AI-generated outputs.

You'll be immersed in:

• Iterating to craft the optimal set of instructions (prompts) for reliable, high-quality AI outputs
• Evaluating to systematically assess results, identify issues, and determine your next iteration

There's no shortcut. This evaluation ↔ iteration loop is an endless cycle, playing a crucial role at every stage of your feature's lifecycle—from discovery and prototyping to delivery and monitoring.

3️⃣ Building AI feature is a TEAM job