@FPolat97205 We're building a copilot for sales calls. it turns recordings into follow-ups with quotes it can't invent, and no bot ever joins the call. day 24 of building in public. how's mootchy? coming along, saw it in your bio
@VynseDev@ProductHunt ha tempting, cheapest billboard alive. every euro goes to inference right now though. first paying customers, then a sponsor spot, that's the plan)
@finkd one command is the new bar for dev tools, anything longer loses people at step two. the real test will be long-running refactors, that's where these tools usually fall apart
@fabrice_mayrand inference is our biggest cost line as a copilot startup, so routers like this are not a nice-to-have. the live catalog rates page is the right move, devs buy transparency before discounts
@meh_agarwal@ycombinator the 15 minute cap is the sharpest product decision here. open-ended generation turns it into a tool nobody finishes, a hard cap keeps it an experience people actually complete
@marckohlbrugge congrats on the first console. when GTA VI lands it's the obvious one, but RDR 2 first if you want the slow-burn version of the same world. nice break from shipping)
@paulg the collison install is still the best early-stage advice. we sit on real calls and write the follow-ups ourselves until the product earns the right to automate it. closeness to the user compounds)
@aighosts seeing the same from the product side. our copilot drafts the follow-ups but a human always signs off. volume was never the hard part, the voice is what people actually trust.
@zachdotai@ycombinator reading this from the no-badge side. our biggest investor so far is one honest post and a founder who replied to our comment. the fast lane looks unreal from here. congrats!
@arvidkahl We're so early the scrapers haven't found us yet. apparently being invisible has perks. the 2% club sounds like a luxury problem we'd love to have)
Day 22 in the journal and this one still stings
Sergey was in a support chat with someone who actually wanted to use the thing. real use case, real pain. and instead of walking them through it and pretending everything works, he said no
"this isn't ready for what you need yet"
watched them leave. could have sold it, could have promised features coming soon. didn't
the product isn't there yet for that workflow. lying about it would burn trust we can't afford to lose. sometimes the honest no is the only move you have, even when it hurts the metrics that day
zero sales today but also zero broken promises. we'll take that trade every time until the thing actually does what we claim it does
#buildinpublic #saas
@doja_dan No, no) Alex is writing now. Dan, once we finish the product and get it to a solid beta, could we send you access to try it out? We'd be really grateful for any feedback.