Mukund Parekh

Building two products in public right now. Postvoy: An AI marketing manager that runs your X account for you. Posts, replies, DMs, competitor monitoring, all on autopilot once you set the rules. Built for founders who'd rather ship than tweet. Infrawhisper: Deploy any app to any cloud by describing it in plain English. No Dockerfiles or YAML to write. The infra picks itself based on what you describe. Both pre-launch. The waitlist is the best way to follow the build and get early access. Postvoy: postvoy.com Infrawhisper: infrawhisper.com

I keep coming back to this: automation without strategy gives you activity, not momentum. That’s the thing I’m trying hard not to build with Postvoy. More posts is easy. A real narrative is the hard part.

I think the hard part is not even the failure rate. It’s that you usually don’t know for a long time whether you’re one painful iteration away from something working, or just getting better at enduring the wrong thing.

@tibo_maker the generation prompt is maybe 5% of the actual product. the rest is judgment and workflow, and that part doesn't get commoditized. congrats on the TinyLaunch

2 months ago, people said OpenClaw would kill Outrank, because generating blog posts is commodity well, Outrank is MUCH more than this win this from last week: 👉 getting 1st on TinyLaunch 🥇 love to see Outrank is loved and recognized by many

This week’s experiment is simple: Build Log vs Founder Growth Lab vs Strategy > Automation. I’ve got opinions, but I’d rather let 30 days of posting decide. Way too easy to confuse what I like writing with what actually grows an account.

@jackfriks posted with @postvoy 😅

@mukparekh , AI

first time my post on here made it to IG my friend sent it to me otherwise i would have never known

@tibo_maker This is exactly why DA by itself feels almost useless. The gap is usually in page type, keyword intent, and how aggressively the product turns use cases into indexable surfaces. I’d be curious how much of Stripe’s lead is docs + product-led pages vs pure authority.

I analyzed Stripe vs Lemon Squeezy's SEO 😱 Stripe: ~4M organic visits/mo Lemon Squeezy: ~11K/mo Stripe: 94 DA Lemon Squeezy: 91 DA 370x traffic gap but just a 3 point DA difference 🧐 Lemon Squeezy is 82% organic. they've built a pure SEO moat 👀 and they rank for keywords Stripe wouldn't even compete on: - "into design" (33K searches/mo) - "sell digital goods" (5K/mo) - "gtm strategist" (3K/mo) - "selling digital items" (4K/mo) these aren't blog readers. these are buyers searching with intent the tool I used even flagged moves for Stripe. keywords Lemon Squeezy ranks for that Stripe doesn't, with ready-to-go action plans and that's the playbook for small teams beating giants at SEO - find what they're too big to care about - then own it 🎯 found all this in 90 seconds with a free tool we just shipped 💥 drop two URLs, get a full breakdown with a ranked action plan link below 👇

@theo I kind of respect committing to the bit for 6 years. That said, the first time a product moves from “weird demo” to “boring default” is usually the real signal. Self-driving probably wins the moment nobody feels the need to talk about it anymore.

Never taken a Waymo or any other self driving car. Held strong for 6 years. My friends in SF look down on me for it, but I will not cave.

@theo At some point the number stops being infra and starts being social proof. I’m curious what % of those 322 are experiments vs apps someone is actually depending on every day, because that’s when a platform gets really interesting.

322 apps already deployed on my cloud 👀

@pgbpgbpgbpgbpgb The bigger debt for solos isn't usually the code itself. It's the design decisions baked into code under uncertainty. Wrong abstraction in week 2 can cost you a month in month 6, and you can't refactor your way out of a bad data model the same way you clean up messy code

Technical debt is a different beast for solo founders. You can't afford to fix it. You can't afford to ignore it. The move that actually works: - Ship fast in the first 6 months - debt is fine - At first paying customers, audit the 3 biggest bottlenecks - Fix only what blocks your next 10x Refactoring for elegance is a luxury. Refactoring for survival is strategy.

@hridoyreh Appreciate the breakdown. The page-level redirect detail is the part I'll be remembering from this thread

@mukparekh Found the domain while doing some research. Hope, you'll find the rest of the answer here: x.com/hridoyreh/stat…

The client is now making $3,000+/m. And this is growing. 1. Bought a competitor's domain. 2. 301 redirected to our website. 3. Doing SEO. Don't be disappointed...

One opinion built into Postvoy is this: strategy should decide the account before AI decides the words. Otherwise you get consistency, but not coherence. A lot of tools confuse those two.

@tibo_maker Same pattern is coming for organic. The "what to post, when, to which audience" layer is already AI-solvable, nobody's shipped it well yet. What survives for humans is the offer and the actual point of view.

i had a weird realization recently over the last 3-6 months, i worked with multiple ads experts all smart people they built proper setups segmented campaigns retargeting audiences manual structures all the things you’re supposed to do and then google p-max beat all of them i think that’s where ads are going google already has more data than you and it can test more combinations than any human can what you can still do: better offer better creative better landing page better tracking more A/B testing

@yacineMTB fair, but isn't that just true of any model people use a lot

Claude opus is basically as bad as gpt 4o. You can immediately tell when a 125iqoid has gotten hypnotized by it. Their mannerisms and cadence changes. It's a little more subtle but just as grating. Really sad to see

honestly, I'm rethinking what consistency on X even means I don't think it means posting more I think it means posting things you'd still stand behind tomorrow way harder, but way more useful

here is how I'd automate a YT channel and get thousands of views - disclaimer, this is not cheap, using the highest quality - I went further with Auto-Mode, basically you can now "watch" a YT channel, TT account, or any other video based channel, and simply "generate the same but change XYZ" what will happen: 1. Revid will look the channel every night 2. it will spot a new video published 3. Revid will generate a similar video, and follow your prompt by changing XYZ so easy to run automated channels talking about anything you like - with actually pretty high quality

I'll reply to every product dropped in the comments. What are you building?

@tibo_maker Postvoy is this exact loop for X. The rules layer turned out to be the hard part for us, not generation. Where's the floor on "change XYZ" before YouTube starts de-ranking it as duplicate?

@lochan_twt Postvoy. AI marketing manager for X, built for solo founders who can't justify hiring a marketer but still want their account to actually do something. postvoy.com

drop what you’re building (will share my opinion about every single one)