This product was not featured by Product Hunt yet.
It will not be visible on their landing page and won't be ranked (cannot win product of the day regardless of upvotes).

Product Thumbnail

IRA

An offline voice AI that lives on your desktop

Productivity
Open Source
Artificial Intelligence
GitHub
Visit WebsiteSee on Product HuntGithub

Hunted byPAGIDI MAITHRIPAGIDI MAITHRI

Most to-do list apps are just another screen competing for your attention. Task Agent takes a different approach — a small AI orb that lives quietly on your desktop and only speaks when you ask it to. Say a custom wake phrase, ask what you've missed or what's due today, and it replies out loud — reasoning over your real task list using a local LLM. Speech recognition, task logic, and AI reasoning all run on your machine. No cloud, no API keys, nothing leaves your device.

Top comment

Hey everyone 👋 I'm a 2nd year CS student, and I built this because I kept forgetting what I'd actually finished each day — DSA problems, project work, interview prep, learning new frameworks. I wanted something that would just tell me, out loud, without opening another app to check. The build took a lot of dead ends worth mentioning: Chrome's Web Speech API doesn't work reliably inside Electron (Chromium lacks Chrome's private API key), so I switched to Vosk for offline speech recognition. Then discovered the popular Node.js Vosk binding is broken on modern Windows, so I ended up bridging to Python's official Vosk package via a subprocess instead. Packaging it as a real installer surfaced a whole new set of issues around file paths and permissions that never show up in dev mode. Tech stack: Node.js + Express backend, Electron for the transparent always-on-top orb UI, Python + Vosk for offline speech-to-text, and Ollama (Llama 3.2) for local LLM reasoning over the task data. Everything runs 100% offline — no API keys, no cloud costs, no data leaving your machine. Setup takes about 5 minutes after installing (Ollama + a speech model, both free) — full step-by-step instructions in the README. Would love feedback, and happy to go deep on the architecture with anyone curious about the debugging journey!

Comment highlights

finally a to-do app that does not demand my attention every five minutes, the desktop orb idea is genuinely clever and the local-only approach is a nice bonus.

About IRA on Product Hunt

An offline voice AI that lives on your desktop

IRA was submitted on Product Hunt and earned 0 upvotes and 2 comments, placing #124 on the daily leaderboard. Most to-do list apps are just another screen competing for your attention. Task Agent takes a different approach — a small AI orb that lives quietly on your desktop and only speaks when you ask it to. Say a custom wake phrase, ask what you've missed or what's due today, and it replies out loud — reasoning over your real task list using a local LLM. Speech recognition, task logic, and AI reasoning all run on your machine. No cloud, no API keys, nothing leaves your device.

IRA was featured in Productivity (655.7k followers), Open Source (68.6k followers), Artificial Intelligence (473.1k followers) and GitHub (41.3k followers) on Product Hunt. Together, these topics include over 287.6k products, making this a competitive space to launch in.

Who hunted IRA?

IRA was hunted by PAGIDI MAITHRI. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.

Want to see how IRA stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.