Postessia just went live on Peerlist Launchpad — building this between nursing shifts and exams
Quick one — Postessia (AI that writes LinkedIn posts that actually sound like you, not like AI) just launched on Peerlist Launchpad.
For context on why this exists: I kept seeing the same complaint everywhere — founders trying AI writing tools, getting generic output back, and giving up on LinkedIn entirely. Not a time problem. A "this doesn't sound like me" problem. That's the whole bet behind Postessia — voice-matching that's actually built to catch and kill the generic-AI-tell patterns, not just spit out a polished paragraph.
I've been building this solo (well, with one technical co-founder) around nursing college — Sem 4, so this has happened in whatever hours exist between hospital shifts and pharmacology exams. If that context is interesting to anyone, happy to answer questions in the comments.
If you want to see what the actual voice-matching output looks like, or have 30 seconds to support the Launchpad listing, would genuinely appreciate it:
https://peerlist.io/postessia/project/join-the-waitlist--postessia--postessia
Would love feedback either way — brutal is fine, this is still early and I'd rather hear the real reaction than a polite one.
"sounds like me" is a harder wall than most voice tools clear. what usually fails is style-transfer wrapping over the same generic scaffold, so sentence rhythm reads right but the tells still leak, uniform sentence length, tidy conclusions, "here's the thing" openers.
does postessia flag structural tells at generation, or is voice matching mostly prompt engineering on top? the tell-detection layer is where the real work sits.
Solid distinction — most tools do stop at style-transfer, and yeah the tells you named (uniform rhythm, tidy wrap-ups, 'here's the thing') are exactly what gives it away.
Right now Postessia's voice-matching is mostly prompt-engineering across a planner→drafter→auditor pipeline — auditor stage does catch some generic patterns, but I won't pretend it's a dedicated structural tell-detector yet. That's actually the next layer I'm thinking about building — explicit checks for sentence-length variance, ending patterns, transition clichés, rather than just hoping the prompt suppresses them.
Curious what tells you've seen most often — sentence rhythm, or something else? Would help me prioritize what the detection layer should catch first
Postessia also went live on SoftRankings, mate! I can see you have a good visitior counts from Pre-Seed teams.
Check it out: https://softrankings.com/products/postessia/analytics
"Sounds like you" is the hard part — most AI writing tools optimize for polish, not voice. Nursing shifts + Peerlist launch is real build-in-public energy.
Honest question on distribution: are you finding LinkedIn creators on LinkedIn only, or also hunting threads where founders say "LinkedIn feels fake / I gave up posting"?
That's usually where the pain language lives before people search for a tool.
I run a scored Reddit digest for founders (find threads, not auto-post). If you're actively hunting ICP conversations now, I can show how I'd map subs/keywords for Postessia on a 10-min call.
"Sorry for the late reply, this got buried — good question though. Right now mostly LinkedIn-native discovery + some manual outreach, haven't systematically hunted Reddit/other threads for 'LinkedIn feels fake' language yet. You're right that's probably where the sharper pain-language lives before someone searches for a tool.
Appreciate the offer, but I'm heads-down on conversion right now, not adding new channels/tools this month. If the digest idea is something I revisit later, I'll reach out