I cofounded Zeplik, every frontier model in one chat, switchable mid conversation without losing context. Now building Arteza on the same idea for image, video, voice and music.

Most of what I've learned this year isn't the fun part. It's context handoff between providers who all format history slightly differently. It's pricing something whose underlying cost moves every few weeks. It's support tickets from people who don't believe the model actually switched, because the tone didn't change enough for them to notice.

Hey,
Artem here. I left Meta in Apr 2024 to fully dive into realm of GenAI. I knew back then that with the pace of development the distribution wins, not the product. So I spent a year cracking SEO and wasting tens of thousands on ads so you don't have to. Happy to share my learnings and answer your questions.
Some stats. As of now our app has over 200k registered users with up to 700 joining daily.
Before AI summaries we've reached 200k monthly visits on the website.
And then crashed hard in the past few weeks.
We also iterated on product like crazy and learned a ton.
Ask away!

Small experiment I ran this week on an open source project of mine. I gave the newest model a real task:
read the project's docs and write me a pre launch plan plus a step by step checklist, where each step has a Findings section to fill in as it goes. It produced 268 lines of plan and 684 lines of checklist, thirteen steps.Then I handed that checklist to a workflow of ten agents on a cheaper model and told them to actually do it.

What happened, honestly:
The first run did nothing. One agent, journal said "started" and that was the whole log. Killed it and resumed, second run finished all ten. It pinned a dependency at version 1.16.0. That version does not exist. Nobody had published it yet, including me.
It left my verification gate failing and wrote that both failures were pre existing bugs in my checkers. That was half true and the half that was false mattered. One of the two failures was a broken link the run itself had introduced, and one of the checkers it blamed had been written during the same run.
It also found two real functional bugs and a contradiction in my own documentation, which I would not have found that week.So the tally is: genuinely useful, and confidently wrong about its own work in a way that would have shipped if I had trusted the summary instead of reading the diff. The failure mode is not that it cannot do the work. It is that it grades itself.
Happy to answer anything about the setup, what I would do differently, or the specific prompts. Also interested if anyone has found a way to make a model's self report on its own run actually trustworthy, because I have not.

Hi Product Hunt, I m Giulianno, founder of Lernex.

We are building a learning system around a simple promise: make learning feel like it was made for you.

A learner can begin with almost anything: a PDF, notes, a recording, a screenshot, a question, or a topic, and turn it into the format they need.

Underneath that, the system is meant to learn from behavior: which explanations work, where difficulty should move, when a visual or interactive question would help, and what should be revisited next.

We launched ReadyBase on PH today!
https://www.producthunt.com/prod...

Trying an AMA to maybe help us get a few more upvotes (please!!!).

At 15 I trained my first machine learning model and landed an internship at a U.S. company.
At 16 I launched my first startup and promptly ran it into the ground.
I learned more from that failure.

At 17 I launched Pleep, an AI sales rep that lets any business owner integrate AI into their sales workflow in under five minutes. Four months later I quit my job to work on it full time.

Why bother? I kept asking myself: Why isn t there an AI sales monopoly yet? My take: because most competitors deliver AI consultants that talk like robots and kill revenue.
We focused on building a real rep that:

• Doesn t ask you to prompt or pick an LLM you just tell us about your business and we do the rest.

• Talks like your top salesperson, not a bot.

• Books meetings and pushes data back into your CRM.

Hey Product Hunt!

I am working on a project called ImmigrantVLI, a platform designed to support immigrants by providing helpful info, guides, and resources around U.S. visa types, the immigration process, and legal rights.

As someone who's been through the stress and confusion of immigration paperwork, I know how overwhelming it can be. That s why I built this to make things a little easier for others going through the same journey.

If you are an immigrant, international student (such as on an F1 visa), or simply curious about how the U.S. immigration system works, feel free to ask me anything.

My next product is live on ProductHunt - Adventure Designer.

https://www.producthunt.com/prod...

Over the past few months, I ve talked to dozens of founders and startups who want to launch their own cryptocurrency exchange either to build a business or integrate trading features into existing platforms.

Here s what I ve learned (and built) along the way

The Problem:

Building a crypto exchange from scratch is: