How to Get Real Feedback on Your AI PM Portfolio
TL;DR
Most aspiring AI PMs build portfolios in private and ship them straight to recruiters. The portfolios that land interviews go through 3-5 rounds of structured feedback first. This guide gives you the five reviewer types you need, the brief that prevents "looks great" non-feedback, and the iteration loop that compresses six months of trial-and-error into six weeks.
Why Most Portfolio Feedback Is Useless
When you ask a friend "what do you think?", they'll say "looks great" — because that's polite, and because they don't know what to look for. Useful feedback comes from specific reviewers asked specific questions. The fix isn't finding nicer reviewers; it's asking better questions.
Reviewer 1: A working AI PM
Tells you whether your case studies pass the smell test. Anchor reviewer; everything else is supplementary.
Reviewer 2: A hiring manager
Tells you whether your portfolio answers their actual screening questions. Bonus if they've hired at companies you're targeting.
Reviewer 3: A senior engineer
Tells you whether your technical reasoning holds up. Catches AI vocabulary mistakes that hiring managers also catch.
Reviewer 4: A peer aspiring AI PM
Tells you whether the portfolio is intelligible to someone at your career stage. The empathy reviewer.
Reviewer 5: A recruiter
Tells you whether the portfolio passes the 60-second skim test. Different audience than hiring managers; both matter.
The Brief That Generates Real Feedback
A 2-paragraph brief in the message turns a polite "looks great" into specific, structural feedback. The brief gives reviewers permission to be honest and tells them what kind of honest is most useful.
Context — who you are and what you're aiming for
"Aspiring AI PM transitioning from data science. Targeting Series B+ AI startups in vertical SaaS. Looking for first AI PM role."
What you're testing this round
"This round, I want to know if the case studies prove I can ship — not whether the design is pretty." Narrows feedback to what matters.
Two specific questions
"1) Does the eval framework section read as real or made up? 2) Would you advance this past phone screen?"
Permission to be brutal
"I want you to be honest, even uncomfortably so. Easier to fix here than after 100 cold applications."
Where to Find Each Reviewer Type
Working AI PMs
Office hours, AI PM Discord, Slack communities, alumni networks of cohort programs. Be specific in the ask: 30 minutes for portfolio review.
Hiring managers
Your existing network first. LinkedIn search by title + AI keywords for 2nd-degree connections. Warm intros only.
Senior engineers
Your current company. Open-source maintainers whose projects you actually use. Pay them with the most concrete questions you have.
Peer reviewers
Cohort programs (paid or free), study groups, Twitter AI PM circles. Trade reviews — your feedback for theirs.
Recruiters
Friendly recruiters from past job searches. Specialized AI PM recruiters on LinkedIn often offer informal feedback if you ask politely.
Get Portfolio Reviews Built Into Your Curriculum
The AI PM Masterclass includes structured portfolio reviews from a Salesforce Sr. Director PM and peer cohort feedback — the five reviewer types in one program.
The Iteration Loop That Compounds
Round 1: Internal pass (week 1)
Self-edit. Read aloud. Cut everything that doesn't earn its space. The cheapest round. Don't skip.
Round 2: Peer review (week 2)
Two aspiring AI PM peers. Trade reviews. Goal: find structural issues you missed.
Round 3: Working AI PM review (week 3)
One person currently shipping AI products. Goal: smell test on the case studies.
Round 4: Hiring manager review (week 4)
Someone who has hired for the role. Goal: would they advance you past phone screen?
Round 5: Recruiter pass (week 5)
60-second skim test. Goal: does the portfolio survive low-attention reading?
When to Stop Iterating and Ship
Iteration without a stop date becomes avoidance. Ship the portfolio when your last two reviewers gave you no structural critiques — only nitpicks. Nitpick rounds compound diminishing returns; structural rounds compound real returns.
Stop signal: same feedback twice
If two reviewers give you the same feedback, it's not a coincidence — fix it. Don't ask a third for confirmation.
Stop signal: nitpicks only
When feedback shifts from "your eval section is weak" to "use a different verb," you've hit diminishing returns.
Stop signal: deadline pressure
If the role you're targeting closes in 2 weeks, ship and adjust live. A portfolio in the wild beats one polished forever.