What are you building? Drop your AI product below

Building TroiAI Inspektor. Chrome extension that puts EU compliance into the ticket where the work happens (Jira, Confluence, GitHub) instead of a portal nobody logs into.

• For: Champions in European mid-market teams (PMs, BAs, engineers) who need compliance answered for the specific ticket in front of them, not the company in the abstract. Project-level compliance, not strategy-level.

• Hardest part right now: bridging two audiences that both have to co-exist for compliance to actually work. Compliance officers and DPOs already know what needs to be done. Champions don't, and increasingly can't afford not to. Our real job is translating expert knowledge into the champion's next specific action for the ticket they're working on. Get it wrong and compliance ends up audit-ready on paper or blocking the sprint. Get it right and GDPR or CRA actually protect the data and systems they're meant to.

Live in the Chrome Web Store today: https://chromewebstore.google.com/detail/trusttroiai-inspektor/coibbfeednldijipnmkcdmapopelblog
Full Product Hunt launch coming in September.

Engineers on my team started shipping 40x faster once they picked up AI coding agents. I'm the PM — I already didn't have enough hours in the day, and once that happened the bottleneck flipped onto me: my job was never writing code, it's making sure they build the right thing, and there was no way I could keep up with that pace by hand. There's a ton of tooling thought through for the engineer side of this shift. Almost nothing built for the PM side — the surface to actually feed an agent what matters just didn't exist.

So I built Atisbo to solve that for myself. I'm a PM and I was tired of the operational work of maintaining a backlog by hand — I built something to do it for me, so my agent could work alongside me instead of me being the thing slowing it down. It turned into a context layer: user feedback, sales calls, support tickets all get synthesized and grouped by actual root problem, not by whoever filed the ticket first. Instead of a doc I have to keep rewriting, it's built to be read directly by coding agents over MCP.

Who it's for: PMs already shipping with AI coding agents (Claude Code, Cursor) who are still doing Product Ops by hand — rebuilding "what's actually going on" from scratch before every decision.

The number that convinced me to keep building it: it cut about 18 hours a week of that reconstruction work off my own plate. And watching teams that pair a PM with a strong product culture, decisions and shipped fixes moved roughly 7x faster than the same PM working alone with Claude Code and no context layer feeding it — same agent, same PM skill, just missing the "why" behind what to build.

Hardest part right now: two things, actually. First, getting the grouping right without a human babysitting it. Clustering similar complaints is easy — not merging two different root causes into one bucket, or splitting one real problem into five duplicates, isn't. Similarity search alone kept silently corrupting the ranking, so we added an LLM-judge pass that has to confirm "same root cause" before any merge is allowed.

Second, making it technically viable, cost-wise. Any expensive model can do this job today — throw a big API call and a good prompt at it and it works. What it won't do is stay sustainable or operable once you're running it against a real customer's volume of evidence, every day, at a price that isn't insane. Finding the actual point where this could run at scale took a while — funny enough, right around the week I landed on it, Fable 5 came out.

Curious if anyone else here has hit that same wall feeding evidence into an agent's context — did embeddings alone hold up for you, or did you also end up needing a judge step?

Building an AI-native LaTeX editor with 2x faster compile speeds and unlimited free collaboration for researchers and students.
https://www.producthunt.com/products/murfy?launch=murfy-ai

Really appreciate any feedback or support!

What I'm building: DriftDetector (https://drift.reweaver.ai) — free, no signup. Paste a GitHub repo, get a drift score in ~1 minute: how far your vibecoded repo is from production-ready, across nine dimensions (security, a11y, reliability, AI code governance...), down to the file/line and the commit that opened each gap. Deterministic, no LLM calls; private repos scanned in memory, nothing kept.

Who it's for: anyone shipping AI-assisted code faster than they can review it.

Current challenge: figuring out what data people actually want in the report. Run it on a repo and tell me what's useful and what's missing — "non-problem" is a valid answer.

Cheers,
Jonathan

What I'm building: Syngulr(www.syngulr.ai) that is to help build ai agents for solopreneurs and small to medium sized businesses that automates day-to-day operations while the AI CEO brain keeps learning from all prompts and inputs to make it so the AI Agents keep learning and improving instead of the same work. It is designed to improve upon AI agents and make them more intelligent and save people more time and money.

Who it's for: It is for solopreneurs, small and medium sized businesses who want to automate and streamline more work with integrations, talk on slack, whatsapp, teams and be on the go and be able to work productively, efficiently and stay lean.

Hardest part right now: Getting the UX and UI right to my standards and then pushing marketing angle. The marketing angle of a tool that can do anything can be hard to identify.

I kept seeing people ask some version of: “What side hustle should I start based on the skills I already have?”

The problem wasn't a lack of business idea lists — there are thousands of those. The problem was figuring out which ideas actually fit you.

So I built a tool that starts with the person's skills, experience and interests, then generates personalized business ideas and a short validation plan.

I'm testing whether personalization makes the result useful enough that people are willing to pay for it.

Building Maleu — India's first life OS 🇮🇳

Social sharing + structured learning (Bloom) + fitness tracking (Wrex) + communities (Circles) + AI camera direction.

All in one app. Free. Launching on PH today: producthunt.com/products/maleu

What I’m building:

I’m building a small validation assistant for early indie founders.

It’s for people who have already built a landing page, demo, or MVP, but still feel unsure what to do next.

The tool looks at the founder’s current product, what they’ve already tried, and the signals they’ve received so far. Then it gives them a clear read on where they are stuck and one small validation action they can do next.

Who it’s for:

Early solo founders, AI coding beginners, and side-project builders who already have a product direction, but aren’t sure whether to keep building, keep researching, or show it to more real users.

It’s not for people looking for a brand-new idea. It’s more for people who already built something and need help making sense of their next move.

The hardest part I’m solving right now:

Finding real early founders with real unfinished validation problems.

A lot of people ask for feedback, but not everyone is actually stuck on validation. Some need growth. Some need positioning. Some just want launch support.

So right now I’m manually reviewing real founder posts and testing whether this can give them a useful next step, not just generic advice.

If anyone here has built something but feels unclear about the next validation step, I’d be happy to try it on a few real cases.

What I'am building:

RunbookAI is an AI-powered incident response platform that detects production incidents from engineering signals (starting with GitHub, with Slack and other tools integrated later), investigates root causes autonomously, generates actionable runbooks, and helps engineering teams resolve incidents faster using AI agents.

Who it's for:

Software engineering teams, DevOps engineers, SREs, platform teams, and engineering organizations that want to reduce incident response time, automate root cause analysis, preserve operational knowledge, and onboard engineers faster.

The hardest part solving right now:

Building a reliable AI investigation engine that can accurately detect incidents from engineering activity, correlate information across multiple sources (GitHub, Slack, CI/CD, logs, etc.), determine probable root causes with minimal hallucination, and generate trustworthy, executable runbooks that engineers can confidently use during production incidents.

Building: adextract - one MCP server that connects AI agents to ad libraries (Meta, TikTok, LinkedIn, and more)

For: Claude power users and AI agent builders who need programmatic access to competitor ad data

Hardest part right now: Every ad library is a different silo with no API. The scrapers that exist break on every DOM change. Making a single endpoint that survives layout shifts across four platforms while returning clean structured JSON — that's the real problem.

PH Listing - https://www.producthunt.com/products/adextract?utm_source=other&utm_medium=social