I'm a program manager at a tech company wanting to upskill and learn more about foundational AI concepts as my company begins to pivot towards AI. I want to understand fundamental concepts as well as how to use AI in my everyday workflow. I learn best through real life examples and case studies.
Welcome! I am excited to work with you on building out your AI product management skills.
Before we dive into the core concepts, I would love to know a bit more about your background so I can tailor our real-world examples.
To help me tailor our examples, could you share a bit about what kind of products your company currently manages, or a specific area of your workflow where you are most eager to start using AI?
We manage enterprise B2B software, and I want to automate requirements gathering.
It's great to meet you! Working on enterprise B2B software while looking to automate requirements gathering is a perfect real-world foundation for this session.
The Core AI Toolkit in B2B
To automate requirements gathering effectively, it helps to break down what different types of AI actually do. In enterprise B2B, three main building blocks come into play: (ML), (NLP), and (GenAI).
Imagine you have 50 transcript hours from B2B customer calls. NLP extracts the underlying pain points, ML flags which feature requests come from your highest-value accounts, and Generative AI drafts the initial user stories for your team.
Make me a diagram of how these three AI types handle customer feedback
Here is a breakdown of how NLP, ML, and Generative AI work together in a pipeline to transform raw customer feedback into structured product requirements.
A three-step customer feedback pipeline showing sequential processing of call recordings into structured user stories.
I have created a diagram showing how these three components pass data along the chain. NLP handles the unstructured text, ML applies business logic to prioritize, and Generative AI turns it into actionable work items.
Make me a quiz on these 3 AI types
Let's check your understanding of these three core AI building blocks before we talk about adopting an for requirements.
Imagine your team receives hundreds of raw customer support tickets written in natural language. You want a tool to automatically read those tickets and write structured user stories from them. Which type of AI is specifically responsible for generating that new text draft?
Machine Learning
Not quite, but that's a very common mix-up! Standard is incredible at pattern recognition and prediction (like scoring which support tickets are highest priority), but it doesn't author new text. Generative AI is the specific branch built to create brand-new content like text, code, or user story drafts.
Now that we have the core toolkit down, let's explore the —the single most important framework when automating requirements gathering.
An AI-First mindset means starting with the problem, not the technology. AI models bring non-deterministic behavior, latency, and extra computational costs. Therefore, a great product manager only applies AI when it provides clear, measurable value that traditional, rules-based software cannot deliver.
For example, if you just want to trigger a task when a client marks a ticket as 'High Priority,' a simple database rule is faster and cheaper. But if you need to read 500 unstructured complaints and draft clean user stories, AI is worth every penny.