I’ll be honest with you — a year ago I was skeptical about AI tools for product managers.

Not because I didn’t believe in AI, but because most “AI for PMs” content felt like lists made for clicks, not for real work. So I did what any PM would do: I tested these tools inside my actual workflow — PRDs, roadmaps, stakeholder updates, research notes, and the messy in-between.

I’m writing this as someone who ships features at work, and also builds on the side — including a platform for gamers and iterate on like a real product (user feedback, content, SEO, continuous improvements).

Here’s what actually stuck.

Why AI tools matter (for PMs)

PM work is full of context switching. One minute you’re in a sprint discussion, then you’re explaining a tradeoff to leadership, then you’re writing a doc that needs to be “clear enough” for engineering and “simple enough” for non-tech stakeholders.

AI helps most when it removes the blank page problem and the “busywork tax”:

  • Turning messy notes into structured docs
  • Summarizing calls and threads
  • Generating first drafts you can improve
  • Speeding up research and comparisons (without you opening 20 tabs)

Used correctly, it doesn’t replace thinking — it buys you time to do more thinking.

My context (so you know where this is coming from)

I’m currently a Product Growth Manager, and I care a lot about building products, not just reporting numbers.

I also recently migrated my personal site (anukulsaini.com) from WordPress to Ghost CMS so I could ship faster, stay lightweight, and get out of the plugin/bloat trap. That migration pushed me into a more “product engineer” mindset: own the full experience, prototype quickly, and optimize end-to-end.

I’m also building an “AI Citations” style tool concept that checks where a website is mentioned across AI search assistants and gives suggestions to improve visibility. That project alone taught me how powerful AI becomes when you combine product thinking + research + automation.

The AI tools I actually use (and why)

1) ChatGPT — my thinking partner

I use ChatGPT the most, but not as an “auto-writer.” I use it as a thinking partner.

What it’s great at:

  • Turning bullet points into a PRD outline
  • Creating multiple positioning angles for the same feature
  • Writing “first-pass” stakeholder updates I can tighten
  • Stress-testing a roadmap decision by arguing both sides

How I use it: I give it context + constraints + audience. Then I ask for structure first (outline, headings, decision options), and only then ask for a draft.


2) Notion AI — where my docs live

If your docs already live in Notion, Notion AI is an easy win.

What it’s great at:

  • Summarizing messy meeting notes into clear action items
  • Turning scattered thoughts into readable docs
  • Creating reusable templates (specs, retros, weekly updates)

It’s not magic — but it reduces the friction of documentation, which means you actually do it.


3) Perplexity — fast research with sources

When I need to research something quickly and I want citations/sources I can cross-check, I use Perplexity.

What it’s great at:

  • Competitive research
  • Market scanning
  • Finding recent references without drowning in tabs
  • Getting a quick “what’s the landscape?” answer before I go deep

This is especially useful when you’re writing a strategy doc and you need external context without guessing.


4) Gamma — decks without wasting your life

I used to lose hours making decks. Gamma makes it fast to go from outline → clean presentation.

What it’s great at:

  • Roadmap/strategy decks
  • Stakeholder updates
  • “Here’s the plan” narratives

It’s not perfect for pixel-level control, but it’s perfect for speed.


5) Dovetail — turning interviews into patterns

If you do user interviews (or even internal stakeholder interviews), Dovetail can save you a lot of manual synthesis.

What it’s great at:

  • Finding repeated themes
  • Grouping feedback
  • Summarizing what changed across conversations

PMs don’t lose time in interviews — they lose time after interviews, when insights stay trapped in notes. This helps unlock that.


6) Zapier (or automation in general) — the glue

Automation isn’t flashy, but it’s the difference between “we should do this” and “we actually do this.”

What it’s great at:

  • Sending feedback to a central place
  • Creating tasks automatically from forms/emails
  • Keeping your workflow consistent

If you feel like you’re “copy-pasting your job” between tools, fix that first.


What I tried and dropped

Not everything is worth it.

  • Tools that force a full workflow change just to use the AI feature (high switching cost).
  • “PRD generator” tools that look nice but don’t beat a strong ChatGPT prompt + your own context.
  • Anything that adds another subscription but doesn’t clearly save time every week.

My bias is toward a lean stack: fewer tools, deeper usage.


How to adopt AI without overwhelming yourself

Here’s the approach that worked for me:

  1. Pick one workflow (PRDs, research, meeting summaries).
  2. Use one tool for 2 weeks in that workflow only.
  3. Track whether it actually saved time or improved quality.
  4. Keep what works; drop what doesn’t.
  5. Only then add the next tool.

The goal isn’t “use AI.” The goal is “ship faster with less stress.”


The bigger shift: PMs who build will win

In 2026, the most valuable PMs aren’t just great at writing docs. They can prototype, automate, and ship MVP-level output faster.

That doesn’t mean everyone must become a full engineer. But having a builder mindset — and using AI to close gaps — is becoming a real advantage.

That’s the direction I’m personally leaning into: product thinking + execution speed + enough technical depth to move without waiting on a full team.


Bottom line

If you’re a PM and you feel like your days disappear into docs, meetings, and “alignment,” AI tools can give you time back.

Start with one:

  • ChatGPT for docs and thinking
  • Perplexity for research
  • Notion AI for writing and structure
  • Gamma for decks
  • Dovetail for synthesis
  • Automation to remove repetitive ops

Use them to remove friction, not to outsource judgment.