I m building LibertyFile One for a specific problem: ordinary U.S. freelancers and small businesses often reach tax season with CSV/OFX/QFX files from several banks, cards, processors, and marketplaces but no clean handoff for an accountant.
The app works locally in the browser or Windows app. It normalizes files, deduplicates rows, conservatively matches equal-and-opposite internal transfers, sends low-confidence items to manual review, and produces an accountant-ready ZIP plus a 2025 filing handoff for all 50 states + DC. SSNs and statement rows are not uploaded.
The core version is free. I d value feedback on the review flow, especially whether the exceptions list makes the unknowns obvious enough at normal zoom (14px minimum): https://tax.507643.xyz/
I m building TransferIQ, a tool that compares traditional remittance providers, banks, crypto exchanges, P2P routes, and stablecoin-based transfers using the final amount received not just the advertised fee.
The core comparison tool works, but I ran into a product problem:
International transfers are low-frequency.
Even if users find the comparison useful, they may not have a reason to return until their next transfer.
Hey Product Hunt I built Kinetic Override for Android.
It's a free, no-root auto clicker + macro recorder for Android 15+. Instead of only tapping one point, it records real gestures: taps, swipes and long-presses, then replays the sequence in loops with timing controls.
Over the past few weeks, I built a reproducible benchmark to explore a practical question:
Can a local language model correct writing effectively while also explaining the user s mistakes?
This work is part of LinguaPilot AI, a Windows writing coach designed to improve the user s own draft and explain the corrections, rather than automatically writing everything on their behalf.
I wanted to share 4 production-ready AI products I ve built recently, exploring different AI architecture patterns from in-browser WASM databases to ATS parsers:
SetFlow AI Smart AI workout generator with quick-commerce price comparisons. Live App: https://setflow-ai-app.vercel.app
Nexis Insights Product feedback analyzer featuring a 128-dimensional WASM SQLite RAG engine for semantic vector search & grounded citations. Live App: https://nexis-insights-app.verce...
Hey Product Hunt! I m Aaron, the solo maker behind Timevora an emotionally intelligent AI life simulator launching soon. Like many people, I used to replay moments in my head wondering: What if I had done things differently? That question stuck with me for a long time. So I built Timevora a simple, private space where you can explore alternate life paths through thoughtful AI-generated timelines. What Timevora focuses on: grounded, realistic scenarios emotional clarity and reflection completely free access privacy-first experience I d love to learn from this community before launch What kind of what if scenario would you personally want to explore? Your feedback will directly shape the next version of Timevora.
We've been building Superbrain, an AI coding assistant . Most coding assistants start losing context as projects grow. We wanted to solve that by focusing on deep context understanding, so Superbrain can handle large codebases and long-running engineering tasks without constantly forgetting previous work.
Hi, I'm Tyler. I built Verlox because terminals ask for blind trust. You run a command, text scrolls by, and you're expected to know if it worked, what it changed, and how to undo it.
Verlox is an AI terminal built around the opposite idea: you should always know what just happened. Every command becomes a card with the real output inside, a real terminal, so colored diffs, progress bars, even arrow-key menus work right in the card. Under each card, plain English: what ran, whether it truly succeeded (real exit codes, not guesses), and what to do next. When a command asks a question, you get labeled buttons instead of a cryptic Y/A/N/L/S row. When you delete something, it goes to the Recycle Bin, recoverable. When you press Stop, it stops.
