Launching today
Parrot
Smart Meeting Assistant for Mac that helps during the call
36 followers
Smart Meeting Assistant for Mac that helps during the call
36 followers
Parrot is a free, open-source meeting recorder for Mac that helps during the call, not after it. No bot joins: it records what your Mac hears and transcribes on device. When a client asks something your pricing sheet or FAQ covers, the answer appears with the file it came from. After the call you get a report of promises made and questions missed, and you can chat with all your meetings. Use Claude, a free local model with Ollama, or your own server. No account, no subscription.
Products used by Parrot
Explore the tech stack and tools that power Parrot. See what products Parrot uses for development, design, marketing, analytics, and more.
LLMs 2
LLMs 2
OllamaThe easiest way to run large language models locally
5.0 (47 reviews)
Ollama is what lets Parrot run with the Wi-Fi off. Pick a local model and the live assistant and the post-call report both run entirely on your Mac, so nothing from the call leaves the machine and it costs nothing. For people who can't send client calls to any cloud, it's the reason Parrot works at all.
Claude by AnthropicA family of foundational AI models
5.0 (1K reviews)
Claude did two jobs on Parrot. It was my coding partner for the whole app, about 24k lines of native Swift that I couldn't have written alone at night after work. And it's the brain behind the best live cards: when a client asks something mid-call, Claude writes the answer from your own documents in a couple of seconds.
Engineering & Development 4
Engineering & Development 4
GitHubHow people build software
5.0 (730 reviews)
Parrot is open source and lives on GitHub: the code, every release, the issues and the automatic update feed. It's also how people check that an app promising "nothing leaves your Mac" really does what it says.
Claude CodeAnthropic’s deep-context AI coder
5.0 (794 reviews)
Parrot was built in Claude Code, mostly late at night after work. I'm not a full-time developer, and it let me ship a native Swift app with 400+ commits and 25 releases. A FILEMAP with one line per file plus a rule to read it first made it fast on a big codebase.
AI Agents 1
AI Agents 1
