Remember when meetings were actually about conversation? When you could look people in the eye, listen to what they were saying, and be fully present? Before everyone was heads-down, typing notes, half-paying attention?
I missed that. A lot. I take a ton of calls for work, and I found myself constantly torn between being in the conversation and capturing what was said. You can't really do both. Not well, anyway.
So I looked at the existing tools. Otter. Fireflies. Granola. They all have the same basic promise: we'll take notes so you don't have to. Great. But then I noticed the patterns.
The things that bothered me
First, the bots. Nothing kills the vibe of a meeting faster than "Notetaker Bot has joined the meeting." It's awkward. Some clients ask about it. Some companies don't allow recording at all. The bot announces to everyone that you're recording, which changes the dynamic of the conversation.
Second, the subscriptions. $20/month here, $15/month there. It adds up. I was already paying for enough SaaS tools. Did I really need another monthly expense for something as basic as "remember what happened in my meeting"?
Third, and this is the big one: the data. Where does it go? Every one of these tools processes your audio on their servers. They have your meeting recordings. Your conversations with clients. Your internal strategy discussions. All sitting on someone else's infrastructure.
Sure, they all have privacy policies that say they won't look at your data. But they still have it. They could look at it. Someone at the company could look at it. A hacker could get it. A government could subpoena it.
"Can't" is different from "won't"
This is the core idea behind Trilla. I didn't want to build another tool that promises not to look at your data. I wanted to build one that literally can't see it.
When you run Trilla in local mode with Ollama or LMStudio, nothing leaves your machine. The transcription happens locally. The AI processing happens locally. There's no server to send data to. There's no account to create. There's no "cloud" involved.
If you want to use cloud APIs (OpenAI, Anthropic, Deepgram), you can. But you use your own API keys. The data goes directly from your computer to the provider. Trilla never sees it, never touches it, never stores it.
I can't see your meetings because there's nowhere for me to see them. The architecture makes it impossible.
Why not just raise money and build a "real" company?
People ask me this. Some of the tools I mentioned have raised millions. Granola just raised $43 million in their Series B. Doesn't that mean they can build something better?
Maybe. But that money comes with strings. Investors want returns. That means growth. That means monetization. That means finding ways to make money from users.
When you have a huge dataset of meeting recordings, there are lots of ways to monetize it. Training AI models. Selling insights. "Improving the product" in ways that benefit the company more than the user.
I don't have investors. I don't have that pressure. I built Trilla because I wanted it for myself. I use it every day. If other people find it useful, great. If not, I still have the tool I wanted.
What's next
I just shipped speaker recognition. Trilla can now identify who's speaking and remember voices across meetings. Name someone once, and they're automatically recognized in future calls. And the voice prints? Stored locally. On your device. Not uploaded anywhere.
There's more I want to build. Calendar integration so you don't have to remember to start recording. Better search across all your meetings. Maybe even a mobile companion app.
But the core principle won't change. Your data is yours. I can't see it, and I don't want to.
If that sounds useful to you, give Trilla a try. It's free. No catch.