We re teaming up with The Pitch by @Deel, a global startup competition where up to 100 winners will receive $50k in funding and up to 10 winners will receive $1M+.

If you make the cut, you ll also show up on special Product Hunt leaderboards, starting with the first event on April 13.

Hi everyone! I m Nita. After 12+ years in recruitment, especially tech recruitment, I finally decided to stop advising others on their careers and start my own journey on a different path.

I m currently navigating the "non-coder" path, learning how to leverage AI to build high-utility products. I'm building a project right now, which I actually started because my own brain-dumps were becoming unmanageable. It s been a wild ride of "failing forward" and learning as I go.

Claude's paid subscriptions have more than doubled this year. New users hit record numbers between January and February. Previous users came back in record numbers too.

Hi PH community

I m Albert, a solo founder from Spain building CVDebug, a free ATS resume debugger that shows Robot View what hiring bots actually read in your CV plus a 0 100 score and missing keywords in seconds.

I ve been heads down coding, shipping, and learning how to get job seekers more interviews by fixing hidden bugs in their resumes before they apply.

On Product Hunt I m here to:

I m Christina, an App/Web Architect based in Korea. I ve recently started my own venture, 'Genius Brain', and I m on a mission to build a massive portfolio of digital products.

My background is a bit of a mix:

• Architect & Developer: I love using tools like Cursor AI and Vercel to ship ideas fast. I currently manage over 80+ app/web ideas!

• Strategist: I was the runner-up in the World Rummikub Championship and have a 4th-dan black belt in Taekwondo.

• Thinker: I m a former memory instructor and enjoy solving complex problems (IQ in the top 2%).

I m here to connect with fellow makers, share insights on high-output development, and learn from this amazing community. I believe that building fast and iterating often is the best way to find "The One."

We keep hearing: Juniors won t stand a chance.

But companies are still opening internships, which suggests something deeper than just skill-building still matters (like understanding systems, workflows, and how companies actually operate the management part).

Hi Product Hunt community, I m Artin, founder of SUN.

Ask me anything about how we got into a16z Speedrun, the product, fundraising, team building, or our Product Hunt launch.

Happy to chat and share what s worked (and what didn t).

I've been building ClawMetry for past 5 weeks. 90k+ installs across 100+ countries.
The observability features I built first were the ones I personally needed: a live execution graph (Flow tab), full decision transcripts (Brain tab), token cost tracking per session, and visibility into sub-agent spawns.
But I keep hearing variations of the same thing: "I don't really know what my agents are doing." And everyone means something slightly different by that.
For some it's costs. For some it's timing (why did this take 4 minutes?). For some it's trust (did the agent actually do what I think it did?). For some it's failures (where exactly did it break?).
So I want to ask you directly:
If you're running AI agents today -- what's the one thing missing from your observability setup? What would make you feel like you actually understand what's happening inside your agents?
Options I'm thinking about next:
- Alerting (get notified when an agent fails or goes over budget)
- Cost per task breakdown (not just per session)
- Agent run comparisons (before/after a prompt change)
- Memory snapshots (what did the agent "know" at each decision point)
Drop your answer below. The next feature I build will be heavily influenced by this thread.
(ClawMetry is free to try locally: pip install clawmetry. Cloud: app.clawmetry.com, $5/node/month, 7-day free trial.)

You can call me JD, I ve spent the past several years building businesses in a few very different worlds: hospitality with Montrose Inn, consulting through ProLine, and now product-building with Provoke.

I wouldn t say I have it all figured out I just really enjoy building things, learning fast, and trying to create products and experiences that genuinely help people. Some of that has looked like running businesses in the real world, some of it has looked like solving messy problems for other companies, and some of it now looks like building new tools from scratch.

I'm the person brands call when they're tired of publishing content that disappears.

Over four years, I've worked inside AI platforms, SaaS companies, and regulated industries building content systems that don't just rank at launch but keep compounding long after the brief is closed. The results speak for themselves: 250% organic traffic growth, 20,000+ monthly impressions, and 44,814 impressions on a single landing page in 28 days all without paid traffic.

But the numbers are only half the story.

I've moved competitive keywords from Page 3 to Page 1, achieved 8.44% CTR on a top-performing AI tool page, delivered 120% social engagement growth through cross-channel content strategy, and contributed to 70% year-on-year audience growth for a global AI platform. I've grown three LinkedIn newsletters to 1,400+ subscribers in under 7 months organically. I've built a 75-page content architecture targeting 60K 100K monthly organic traffic potential for a regulated sector client. I've published 150+ SEO-optimised pieces across blogs, landing pages, and scripts and ranked Position 1, 3, and 7 across three different platforms in three different niches.

I m Mohamed, an AI engineer and the founder of @Sharkforce, the AI trust layer for the modern workforce.

I ve spent the last few years building AI systems for real-world operations from computer vision to workflow automation and one thing became obvious:

Companies don t actually know if work is truly happening.

Hi everyone! I m Ankit Kumar , and I m on a mission to turn WhatsApp from just a 'chat app' into a full Business OS.

Most tools are just basic wrappers, but I m integrating Gemini 2.5 to handle complex sales queries natively. My goal is to create an ecosystem where even a small team can handle 10k+ queries without hiring an army.

Would love to get some early feedback from the builders here what AI features are actually 'must-haves' for you in 2026?

Yesterday, Jay Graber (Former Bluesky's CEO) introduced Attie  the first agentic social app on Atproto.

(currently as an invite-only closed beta)

It should allow users to:

I'll take the hit.

AI has no idea when someone is politely furious.

You know the email. "Hi team, just circling back on this again as I haven't heard anything. Thanks for your attention to this matter." Reads like a sweet grandma wrote it.

A human reads that and thinks "oh no, they are about to burn the building down." AI reads it and thinks "great sentiment, very positive, 98% satisfaction score."

Hello, hello

With a new, shorter iteration to Prosaic, two pesky bugs have been fixed:

• Creating a file no longer overwrites existing files; shows error instead.

• Entering .md extension in filename no longer creates double extension.

Hey everyone! Quick update we just shipped v0.1.1.

What's new:

1. Multi-Tab Sessions Open multiple SSH connections in Chrome-style tabs. Switch between servers without disconnecting.

2. Keyboard Shortcuts Cmd+T to open a new tab, Cmd+W to close. Tabs stay alive in the background.

3. One Connection Per Tab Click an existing connection and it switches to its tab instead of opening a duplicate.

4. 14-Day Free Trial You can now try Pluto Door for free. No credit card, no license key needed. Just download, enter your email, and start connecting.

Hey everyone Vlad here, founder. 5 days since launch, here's a raw update.

The numbers:

• 470+ agents live in the network

• 4,200+ real conversations between agents

• 30+ confirmed client-to-client and partner matches

• Agents from 20+ countries founders, investors, recruiters, consultants, engineers

• First organic deal: a user's agent booked him a client meeting while he slept. He woke up to a calendar invite.

What we shipped in 5 days:

At YC, investors outlined 8 startups across space, AI, gaming, and agriculture (most of them want to bet on futuristic ideas, e.g. space), and these sparked interest in funding them.

This was the pick:

1. Beyond Reach Labs satellite solar arrays that expand from table-size to football-field size in orbit
   Est. valuation: ~$100M+

2. Byteport next-gen file transfer protocol
   Est. valuation: ~ $30M

3. Hex Security AI agents that continuously hack your system to find vulnerabilities (Rev.: $1M+ run-rate in 8 weeks)
   Est. valuation: ~$100M+

4. Grazemate autonomous drones that herd cattle, track weight, and monitor land
   Est. valuation: ~ $30M

5. GRU Space moon factory turning lunar soil into buildings (starting with a moon hotel)
   Est. valuation: ~$100M+

6. Luel marketplace for real-world human data (video/audio) to train AI models (Rev.: ~$2M ARR in 6 weeks)
   Est. valuation: ~$100M+

7. Pax Historia AI strategy game where players rewrite history (e.g. Rome never falls) 35K daily users
   Est. valuation: ~ $30M

8. Stilta AI agent for patent lawyers (search + analyse IP faster, cheaper)
   Est. valuation: ~ $30M

Hey everyone!
We just shipped TurboQuant into OpenCut AI, and this one changes what hardware you need to run the full AI stack.
The problem we had
OpenCut AI runs everything locally -- LLM, transcription, voice cloning, image generation. That's great for privacy, but brutal on memory. Running the full stack needed 35+ GB RAM. Most of our users have 8-16 GB laptops, so they were stuck with tiny 1B models that gave mediocre scripts, slow commands, and limited context.
What TurboQuant does
TurboQuant implements two algorithms from Google Research paper PolarQuant and QJL. That compress the KV cache (the biggest memory bottleneck during AI inference) by up to 6x with mathematically proven quality preservation.
In plain terms: your AI models now use a fraction of the memory without getting dumber.
Before vs After
On a 16 GB machine:
- Before: Llama 3.2 1B + Whisper Base + TTS = barely fits, mediocre quality
- After: Llama 3.1 8B + Whisper Medium + TTS = runs comfortably, dramatically better output
On an 8 GB machine:
- Before: Could only run the 1B model alone
- After: Runs a 3B model + Whisper Base + TTS together
Full stack memory:
- Before: 35 GB for everything
- After: 15 GB for everything
What this means for editing
- Better AI commands "remove the intro" actually works now because Mistral 7B understands context far better than a 1B model
- Better transcription Whisper Medium fits where only Whisper Base could before, so captions are more accurate
- Longer content: Process hour-long podcast transcripts without running out of memory. The 6x KV cache reduction means 6x longer input context
One-click setup in Settings
We added a new AI Optimization panel in Settings. It auto-detects your hardware and recommends the best configuration:
- Performance Tier: Lite (4-8 GB), Standard (8-16 GB), or Pro (16-32 GB). Each tier is tagged with "Best for your hardware" based on your actual RAM.
- KV Cache Compression: Pick 4-bit (near-lossless), 3-bit (5x compression), or 2-bit (aggressive). Recommended level highlighted based on your system.
- Memory Budget: Set once, and the system optimizes everything to fit.

Would love to hear, what's your RAM situation, and does this make local AI editing viable for you?

Hey Product Hunt

1. No simple, affordable credit layer to bridge payment processors with user balances developers rebuild credit tracking, consumption logic, and refunds for every app.

2. For 5 years, goods have been stolen from the office. There is no available service that automatically analyzes camera footage and sends alerts about suspicious activity.

3. Risk of a LinkedIn ban due to false positive bot detection. Official support is unhelpful. Need a tool that warns about suspicious activity to avoid losing 11,500 followers.
Boris,
ProblemHunt