John_V

Hi HN,

I’m John, founder of PostReach AI. For the last 12 months, my team has been building a conversational agent for social media.

The Beta: We are opening our beta today for the HN community. It is completely free while we gather feedback.

A Note on Access: > Because we are managing heavy GPU loads for the visual analysis and content generation, we are provisioning new accounts in batches. You’ll receive your access email within roughly 30 minutes of signing up. We are doing this to make sure the "Brand Brain" training for each new user doesn't crash the system for everyone else.

The Problem: As founders, we all know we should be active on LinkedIn, X, Facebook and Instagram. But the reality is a fragmented mess, you use ChatGPT for ideas, Canva for design, and Hootsuite for scheduling. It’s a context-switching tax that most small teams can’t afford.

What we built: PostReach isn’t a dashboard, it’s a chat interface. We wanted to see if we could collapse the entire "idea-to-published" stack into a single conversation.

Some technical bits the community might find interesting:

-Brand Ingestion: We don't just prompt a generic LLM. We built a "Brand Brain" that ingests your website and past voice to ensure the AI doesn't sound like a generic bot.

-Visual Analysis: Our Image-to-Post feature uses visual reasoning to understand the "mood" of an image. If you upload a photo of your messy dev desk, it won't just say "Working hard", it understands the hardware and context to write something relatable.

-URL-to-Post: We use a custom summarization pipeline to extract the "hook" from any article or blog post and transform it into platform-specific formats (a technical post for X vs. a professional takeaway for LinkedIn).

So why now? We’ve just moved into beta after aggressive testing. We are looking for our first "real world" cohort of users.

The Ask: We would love for the HN community to break this.

-Does the conversational flow feel faster than a dashboard?

-Is the AI output "cringe" or actually usable for a technical audience?

-What’s missing from your current workflow that we should automate next?

We’re around all day to answer questions about the tech stack, the 12-month dev cycle, or our thoughts on the future of autonomous agents in marketing.

I gave it a try and was able to generate some simple apps. Unfortunately, anything beyond that became too technical, so there's very little value in what I could produce with such tools.

That said, I guess $5k isn't much when we consider how much worse it could have been. There's always something to learn. Are you a technical Founder?

I’m a non-technical founder. For the last 12 months, I’ve been in the trenches building PostReach AI, and I did almost everything wrong at the start.

I’m sharing this because I see a lot of people here about to make the same "outsourcing" mistakes I did.

The $5,000 Agency Trap: I thought I could skip the technical struggle by hiring an agency. I expected a product in 3 months. 5 months later, I had a half-baked, unfinished app and $5,000 less in the bank.

What I learned the hard way:

Co-founders > Freelancers: You need people who are invested in the vision, not just the invoice. My co-founders are the reason my mental health stayed intact when things broke.

Validation isn't "Feeling Confident": I assumed the market was validated. I was wrong. You can't just build and hope they find you.

The Landing Page is a Living Document: We rebuilt ours 4 times. Don’t obsess over it on Day 1. Use it to see if anyone even cares enough to join a waitlist.

Launch Before You’re Ready: The "perfect" app is a myth. Launching early is the only way to get the feedback that actually matters.

Ignore VCs (Unless You Need Them): Don’t waste your limited hours on pitch decks if you can spend them on product.

We eventually found our way and built PostReach AI, a conversational engine that handles social media so founders can get back to building. But the tool is secondary to the lesson: Don’t try to be a solo hero.

I’d love to hear from other non-technical founders: what was your "expensive" lesson? I'll be in the comments to chat.