The AI-Augmented Product Director
A 6-Part Series on What AI Actually Looks Like in Product
Most of what is written about AI and product management is either overhyped: "AI will replace product managers!" Or the opposite: "AI is just a fancy autocomplete."
Neither view is useful if you are a director trying to ship better products faster with the same number of hours in the day.
This series is something different. It is tactical. It is grounded in what I actually do, day in and day out, running a product team at a healthcare technology company. No thought experiments, no imagined futures. Just concrete techniques you can pick up and use this week.
Why I Started Paying Attention
I started paying attention when I looked honestly at where my time was actually going. It was not the hard parts of the job.
The parts where I know exactly what needs to be written, but getting the first draft on paper takes two hours. The parts where I have already synthesized the key insight in my head, but turning it into a structured document for eight stakeholders takes another couple hours. The parts where I am re-reading the same meeting transcript for the third time trying to pull out one decision that was buried in twenty minutes of discussion.
AI did not replace my judgment. It replaced the friction between my judgment and the output.
That is the frame for this entire series.
What This Series Covers
Over the next five posts, I am going to walk through four categories where I have found real value - not just a cool demo:
Post 2: PRDs and Requirements How I have changed the way I write product requirements — from the initial concept all the way through acceptance criteria and cut the time it takes in half without nerfing quality.
Post 3: Stakeholder Communications The email tax is real. Product leaders often spend 2-3 hours a day on communication that, if we are honest, does not require our full cognitive load. Here is how I have reclaimed a huge portion of that time back.
Post 4: Research and Synthesis Competitive analysis, user feedback themes, market positioning; AI has become my research analyst. I will show you what that actually looks like, including where it falls apart.
Post 5: Workflow Automation This is the one I am most excited to talk about. Beyond writing assistance, AI agents can run recurring tasks for you: meeting notes, status updates, ticket grooming, all without you having to think about them.
Post 6: Putting It All Together How the pieces fit into a coherent workflow, what I would do differently if I were starting from scratch, and the honest list of things AI still cannot do.
A Few Ground Rules Before We Start
I am not going to tell you to use any specific tool. The principles here work whether you are using Claude, ChatGPT, Gemini, or something built into the software you are already using. What matters is the approach.
I am going to be specific about the prompts. Vague advice ("just ask AI to help you with your PRD") is not useful. I will show you the actual patterns I use, including the framing and context that makes them work.
I am going to tell you where it does not work. The goal is not to make AI sound magical. Knowing where it breaks down is just as important as knowing where it shines.
The One Mental Shift That Makes Everything Else Work
If I distill everything here into one idea, it is the following:
Stop thinking of AI as a writer. Start thinking of it as a thought partner that never gets tired and never judges you for a rough first pass.
The best way I have found to use AI is not to ask it to write something for you. It is to dump your messy thinking at it — the half-formed idea, the bullet points from the meeting, the three different initial ideas you have been thinking about the problem and ask it to help you find the structure.
You are still doing the thinking. You are just not doing the formatting, the initial drafting, and the synthesis alone.
That is the unlock. It changes how you work with every tool in the toolbox.