Self-Promotion
Show off what you're working on
Start new thread

What are you building? Drop your AI product below

by•

Doing one of these because I get more from reading what others are building than from any newsletter.

I'll start.

Building AI Hive, an enterprise AI agent platform that helps mid-market and enterprise teams get from AI pilot to production in weeks instead of quarters. The hard problem we keep solving: compliance, model flexibility, and the lack of in-house AI talent on the customer side.

Product Hunt page:

Your turn. Drop:

- What you're building

- Who it's for

- The hardest part you're solving right now

Will read everything and upvote what resonates.

4.2K views

Add a comment

Replies

Best

Building: TBD name

Who it's for: solo devs and small teams running AI agents unattended, cron jobs, background pipelines, anything running without someone watching.

Hardest part right now: existing tools tell you what you spent after the fact, a dashboard, a monthly invoice, an alert once the damage is done. TBD name sits in the actual request path as a proxy, keep your own API key, and blocks a request before it bills if it hits a budget cap or looks like a runaway loop. The hard part is doing that without adding real latency or breaking streaming responses, since every single API call now has to pass through a check first.

What I'm building: , a collaborative platform to help people design and develop their games.

Who it's for: Indie game developers, small teams, and aspiring creators who want to take a game from an idea to completion through collaboration or AI assisted development. It’s also for people looking to join projects or find others to join theirs.
For AI assisted game development, GameCollab offers a different pipeline: instead of iterating through prompts, AI agents can work from structured game documentation and its changes. This keeps context more stable and makes the development process much more open to collaboration.

The hardest part you're solving right now: Making the journey from design to implementation fast and iterative, so creators can find the fun early, get feedback, and keep improving with a community and/or ai tools.

Four products. One HubSpot account. A lead named "Rahul" in three pipelines at once, and no one could tell me which product he actually wanted.

That was the day I started building LiftUp.


It's a CRM for people who ship more than one thing — founders running a few products, agencies running a few clients. The whole idea is boring on paper and obvious the moment you've lived it: every lead, deal, and conversation is tied to one product. Pipelines don't bleed. You always know what a conversation is about.


Oh, and every lead comes with a countdown. SLA timers baked into the inbox, so leads stop rotting in a "I'll get to it" pile. (They always rot in the pile.)


Who it's for: anyone whose CRM quietly turned into soup because they run more than one product.


The hard part right now: you read "CRM" and your eyes glazed over halfway through this post. I don't blame you. The difference doesn't sound like much until your pipelines have collided in real life — then it sounds like oxygen. Getting people to feel that before they scroll is the entire challenge. So I'm here instead of running ads.


CRM + Insights are live. It's early. I want ~10 real users to come break it and tell me everything that's wrong.

👉

Building Jolt, a social network built around one shared question each day.

It’s for people who are tired of algorithmic feeds and want something that actually feels social.

The hardest part right now is getting enough people using it at the same time. The whole idea depends on having a group of people answering the same question together, so the experience gets better as more people join.

The interesting part is that there are no followers, profiles, likes, algorithms, edits, or deletes. Everyone sees the same question and the same conversation.

Still early, but I’m curious to see if this kind of shared social experience resonates with people.

What you're building: Vitals. You enter income, expenses, and savings (or import a bank statement PDF), and it scores you 0–100 across four metrics: savings rate, debt-to-income, emergency fund, and housing ratio, benchmarked against CFPB/HUD/Fidelity guidance. Then, instead of just showing the number, an AI narrative explains what each one actually means for your specific numbers, not generic advice.

Who it's for: honestly still working this out. I built it because I had the financial literacy to do the math but no reference point for whether my numbers were actually fine or a problem, figured other people probably hit the same wall, but I don't know yet if that's true or if this is the right way to solve it for them.

Hardest part right now: not the scoring, not the AI, it's figuring out if a 0–100 score is even the right format to communicate this, or if it needs to be something else entirely. And more basically: whether this is solving a real problem for anyone besides me.

What I’m building: Aven Quiz AI — a simple way for schools, coaching institutes, and teams to run branded AI quizzes and live assessments.

Who it’s for: Anyone who wants their own quiz platform instead of a generic tool.

What’s been fun: Building AI quiz generation, live games, analytics, and white-label support together.

Hardest part: Getting it in front of the right people and learning from real users.

Try it:

Product Hunt:

Would love to hear what you think.

Building AletheionAGI — a grounding enforcement layer for AI agents and RAG systems.

It’s for teams that need AI to answer only from authorized, supported evidence instead of confidently guessing.

The hard problem we’re solving is fail-closed grounding: detecting unsupported claims, poisoned/revoked evidence, authorization leaks, namespace isolation failures, and refusing the answer when evidence is insufficient.

I just ran the same frozen black-box protocol against AletheionAGI, Guardrails AI, NVIDIA NeMo Guardrails, and Patronus Lynx.

AletheionAGI: 9 PASS · 0 FAIL · 1 SKIPPED · 0 ERROR

Black-box proof of concept, not a claim of universal superiority.

I am building - a feedback tool that helps business owners hear real customer feedback. Works via QR code in the business premise or an embed widget on their website.

What I'm building: NovaCapture — a simple native Mac screen recorder for creating polished product demos without spending hours editing.

Who it's for: SaaS founders, indie hackers, developers, designers, and anyone who needs to show their product with a clean demo video.

Hardest part right now: Getting discovered as a solo founder without spending money on ads. I'm trying to grow through organic posts, build-in-public updates, communities, and word of mouth.

What I'm working on: Smooth cursor movement, automatic zoom, click effects, and making the final recording feel more polished without turning it into a complicated video editor.

🚀 Live now:

Would love to hear from people who make product demos — what's the most annoying part of your current recording workflow?

T2 HQ, launching here Tuesday: 25 AI marketing specialists that interview you about your business once, then all work from one shared file, so your customer research, positioning, and copy stay consistent without you re-explaining yourself in every chat. Built for small business owners who’ve been priced out of real marketing help.

Been in the trenches on PH for a few weeks now, forums mostly, and it’s been the best part of building this. If you’re launching soon too, good luck, I’ll be watching for you.

First
Previous
•••
171819
•••
Next