On Trust, Privacy, and Ownership

When we talk to our colleagues and friends, we always want to be sure that the things we discuss stay between us. And that's normal. That's how we build trust between people.

The same applies when we interact with our computers. We store personal information, have private conversations, and work with sensitive data. We want to trust that all of it stays under our control.

Recently, you may have seen news about major security breaches and data leaks at large companies. Here are just a few examples:

• OpenCode, October 2026 — A reported security incident related to a large-scale distillation attack affecting OpenCode Console users.
• npm / Keyv & Cacheable, August 2026 — A supply chain attack targeting npm packages and sensitive developer information.
• Salesforce / Salesloft Drift, August 2025 — Attackers used compromised third-party access tokens to steal corporate data.
• Allianz Life, July 2025 — A breach involving a third-party cloud database exposed personal information.
• Ticketmaster, 2024 — Unauthorized access to a cloud database exposed customer information.

Recently, I was reading Lifestreams: A Storage Model for Personal Data, a research paper written by Eric Freeman and David Gelernter in March 1996.

What surprised me was how many of the problems they described are still relevant today. Organizing personal information, summarizing documents, tracking information, personal agents, and making it easier for people to find what they need.

Almost 30 years later, we're still trying to solve many of the same problems, now with AI.

Another paper that inspired me was Local-first software: You own your data, in spite of the cloud by Ink & Switch.

It discusses something I find especially important: why should we have to choose between the convenience of modern cloud applications and owning our own data?

We should be able to work across multiple devices, synchronize information, and collaborate with others without giving up control over our personal data.

These two papers made me think more about what kind of software I want to build.

Building a product that stores a lot of internal communication between people, I believe it's incredibly important to me personally to create a tool that will be a safe place for people.

This means that people should be able to control what data they store, own that data, and not have to rely on centralized storage for their private information.

That's why I'm building a local-first AI meeting notetaker.

Your conversations, transcripts, and notes should belong to you. They should be stored on your own devices, encrypted, and securely synchronized between your devices without a central server being able to read their contents.

Audio should be processed without being permanently stored. The information you decide to keep should remain under your control.

You should be able to use the product offline, decide what information to keep, and stay in control of your own data.

I want to make this tool available on macOS, Windows, and Linux, as well as mobile devices, including iPhone and Android.

I plan to release it step by step, starting with desktop platforms and expanding to mobile devices over time, rather than launching everything at once.

I believe it's possible to build useful AI tools without asking people to give up ownership of their information.

And that's what I want to build.