Seeking PMs for a quick chat: AI PM concept that disagrees with you and remembers your killed bets
Hey everyone,
I’m working on a new AI product-management platform and would love 15 minutes of your time to learn about your daily AI workflows and get unfiltered feedback on my concept.
The Problem Current AI assistants are brilliant amnesiacs and sycophants. They forget your context every session, agree with whatever you say, and do nothing to stress-test your decisions before they ship. Meanwhile, organizations constantly forget why past bets were killed, leading teams to accidentally build the same failed solution twice.
What I'm Building: Arrwin Arrwin is an AI platform focused on judgment-defense.
- It remembers and challenges: It builds a decision graph from your normal usage. If a new recommendation contradicts a past logged decision, Arrwin pushes back and challenges you with a citation (e.g., "This contradicts a decision you killed last month—what’s different now?").
- It stress-tests: Every output is engraved with unvalidated assumptions and specifically calls out who would object to your proposal.
- It earns autonomy: Arrwin joins as a "PM Intern". It earns trust through a silent scorecard of wins and must explicitly request a promotion to gain more autonomy.
Who I'd love to chat with:
- Startup PMs (Doers): Tool-forward PMs who already use AI but lack extra hands for research.
- Enterprise IC PMs (Analysts): PMs in large orgs who are drowning in surface area and lack proactive anomaly flagging.
If you're open to a quick chat to discuss what your current AI gets wrong and how your org handles "killed bets," please drop your AI usages and let's take the discussion from there!