2
6 Comments

Building CognetOS in public, starting today

I run more than one project solo. Most of the time I lose isn't from anything hard - it's from re-deciding the same small things every day: what to prioritize, what to approve, where to even start.

CognetOS is an attempt to fix that specifically. Not a dashboard to check, not another list to maintain - something that makes the call once and tells you what it did.

Still early, still figuring out the exact shape of it. No committee, just building it solo and posting the real process here - what works, what doesn't, what changes.

If you want to see where it ends up: https://getcognetos.netlify.app/

posted toAvatar for product CognetOS
CognetOS
  1. 1
    The interesting test is whether CognetOS makes better calls, not just fewer decisions. What evidence would tell you its recommendations are actually better than the founder’s usual judgment?
    1. 1
      fair challenge, and i don't have that evidence yet. no real usage at this stage to back up "better than my own judgment" what I'd actually want to track once there's real data: not whether its calls beat mine on some abstract scale, but whether the calls I end up overriding or regretting go down over time. If I keep overturning it in the same category, that's a signal it's wrong there, not that I'm just impatient. If overrides stay rare and specific, that's closer to real evidence. But I'll be straight about the actual bet right now - it's not "this decides better than me." It's "re-deciding the same small thing five times a day costs more than one good-enough call made once and left alone." Consistency over brilliance, at least to start. Whether that clears the bar you're describing is genuinely one of the open questions.
      1. 1
        That’s a much clearer product bet — consistency rather than trying to outperform judgment. I’d be interested in seeing what the real usage data says once you have it. If you’re open to it, what’s the best email to reach you on?
        1. 1
          I didn't realize comments don't support links - you can find the link to the wishlist in the description above; thanks.
        2. 1
          Glad that landed clearer. I don't have a dedicated email set up yet - still no domain of my own, running off a netlify subdomain for now. But the waitlist does exactly this: [link], drops you on the list and I'll reach out directly once there's real usage data worth showing, not before. Or just DM me here on IH if you'd rather skip a form it works just as well, probably faster on my end honestly.
          1. 1
            Makes sense. I’ll leave you to collect the usage data and see what the real signal looks like.