Happy New Year! Today marks the first day of a challenge I’ve set for myself: 90 days of content about Lotti1, the app I’ve been building for the past few years.
Rather than explain everything at once, I thought I’d start by showing you. Below is a 10-minute demo of what Lotti can do today.
What you’ll see
The demo walks through task creation and management—the foundation of how I use Lotti daily. It’s not flashy, but it’s the core of what makes the app useful to me: a place to capture what I’m working on, attach context, and come back to it later without losing the thread.
What Lotti is about
Lotti is a personal assistant for tasks, notes, and audio recordings. It’s named after my grandmother, who had an extraordinary memory for the details of people’s lives.
The idea behind Lotti is simple: I want to externalize my memory without handing my thoughts to companies whose incentives don’t align with mine. Everything stays on your devices. When you want AI features—transcription, summaries, asking questions about past work—you connect directly to providers with your own API keys, or run models locally. There’s no Lotti server. I don’t see your data because the architecture doesn’t give me access to it.
It’s open source, GPL-3.0. You can read every line of code on GitHub.
Where to get it
If you’re on Linux, Lotti is available on Flathub today. I use it daily on macOS and iOS via TestFlight—broader access there is coming soon.
What’s next
This is a warming-up post. Tomorrow there will be more. Over the next 90 days, I’ll be sharing demos, setup guides, and the thinking behind how Lotti works. My goal is to find people who find this useful—my tribe, if you will.
If any of this sounds interesting, subscribe to follow along. I’d be glad to have you.
Cheers,
Matthias
The rule I’m following is from James Clear: never miss twice. “The first mistake is never the one that ruins you. It is the spiral of repeated mistakes that follows.” At a minimum, that’s 45 pieces of content over 90 days.
