Andy

Founder building Outcomet
AI-native product operating system
Strategy → Discovery → Capabilities → Feedback
Building in public

[Andy @OutcometApp]

Month 4 as a solo founder. the work moves fast now. what I underestimated is the silence, no team to tell me I'm wrong before the market does. learning to be my own dissent. #buildinpublic #productmanagement

[Andy @OutcometApp]

@AdamRackis agreed. vibe coding nails the idea phase. it's keeping it alive in production people underestimate. solo, I feel both ends, building got easy, running didn't.

[Andy @OutcometApp]

@_catwu this is my favorite kind of AI win, the boring one nobody brags about. flag cleanup is the chore every team postpones forever. did parallelizing it change what you'd trust Claude with next?

[Andy @OutcometApp]

@lennysan The 'want to start their own company' number is the interesting one. Wanting out of the seat isn't the same as wanting the work of building. The gap between those two is where most people stop.

[Andy @OutcometApp]

@shreyas tokenmaxxing might be the perfect word for it :) everyone feels the gap between 'we used AI' and 'this got better', but only one survives a board deck. think the optics phase burns off, or becomes the new normal?

[Andy @OutcometApp]

Velocity is the most overrated metric in product right now. AI gave everyone speed and no one judgment. The teams pulling ahead aren't shipping more, they kill the wrong bets faster. #productmanagement #productstrategy

[Andy @OutcometApp]

Reminder to myself: Always check if your agent uses caching properly.

#buildinpublic #ai

[Andy @OutcometApp]

A feature ships once. A capability ships the next ten. Roadmaps count features, which is why the list grows but the product never gets easier to build on.

#productstrategy #productmanagement

[Andy @OutcometApp]

building got cheap. talking to users didn't. discovery is the part of product work nobody automated, and that's where the real work moved now.

#productmanagement #productdiscovery

[Andy @OutcometApp]

@simonlast This is engineering management with agents instead of people. The teams who already knew how to manage have the head start, not the ones who can prompt.

[Andy @OutcometApp]

@chandrarsrikant Agree, and it's the generalist's moment. Those role splits were always a side effect of tooling, not how the best people actually worked. AI just removed the handoffs that forced everyone into a lane.

[Andy @OutcometApp]

@philipjbeans @lennysan @danshipper This is the fun part of the PM job coming back. I used to wait on eng to spike an idea. Now the prototype is built before the standup ends.

[Andy @OutcometApp]

@damoosmann @lennysan @danshipper Your mum runs semantic dedup. Codex runs string dedup. Most agent bugs live in that gap.

[Andy @OutcometApp]

@andrewchen Above /goal is someone with full conviction and zero evidence. We call it strategy and pay it the most.

[Andy @OutcometApp]

@lennysan @danshipper Bullish on PMs who decide, not the ones who groom a backlog. That half just got cheap.

[Andy @OutcometApp]

planning canvas shipped this week. themes on a graph, drag them into Now/Next/Later, score-ranking does the first pass. took about a month to find the right mental model for it.

https://t.co/oAU7fc2Zol #productops #productmanagement

[Andy @OutcometApp]

Added planned future work to the changes timeline. Closed PRs from GitHub on one end, what's coming on the other. Happy with how it looks already.

#buildinpublic #ux #productmanagement

[Andy @OutcometApp]

Tuning the UX of the capabilites landscape was fun today. See before and after 😃 ... what do you like more?

#buildinpublic #productmanagement #productops

[Andy @OutcometApp]

Playing with new graph visualizations for my feedback->theme->capability relations.

#buildinpublic #productmanagement #productops

[Andy @OutcometApp]

Small paid accounts on X now even have it harder without data.