Parrot

What it does

Parrot records Mac audio during video calls and phone conversations, then transcribes the recordings and surfaces relevant information during the call using an AI assistant. The assistant can reference documents the user has uploaded—like pricing sheets, FAQs, or playbooks—and cite them as sources when suggesting answers. After a call ends, users can ask questions about past recordings. The tool works with any application that produces sound: Zoom, Google Meet, Teams, or speaker phone.

Who it is for

The tool targets people who conduct many calls and need to recall details or prepare responses quickly. Business owners, salespeople, and customer-facing professionals could use it to avoid forgetting commitments or missing opportunities to address objections. The pre-call brief feature and customizable profiles suggest it is designed for sales teams handling different types of calls, though the recording function applies broadly to anyone managing multiple conversations.

Pricing

The site does not show prices.

How it stands out

Parrot emphasizes privacy and local processing. Audio transcription, speaker detection, and document embedding happen on the Mac itself using Apple's language models. Documents are never uploaded to the cloud. The assistant does not join meetings as a visible participant—it only records what the Mac's audio system captures. Users can opt into cloud transcription but do not have to. The tool is open source under GPL-3.0 and requires no account or login to function. It auto-detects language and works across many platforms because it relies on system-level audio capture rather than integrating with specific services.

What a founder should check

First, verify the switching costs and network effects of incumbents. Otter.ai, Fireflies, and Grain have integrated directly with Zoom, Teams, and Google Meet for years, offering cloud storage, team collaboration, and cross-platform access. Parrot's local-only model eliminates these integration points. Understand how much users value those features versus privacy.

Second, test the quality and accuracy of on-device transcription and document embedding using Apple's models. If users expect accuracy comparable to cloud-based services, the local approach may show gaps in noisy environments or with accents and technical vocabulary.

Third, examine the moat around document grounding. Parrot allows users to upload FAQs and playbooks so the assistant references them during calls. Check whether this workflow—managing documents, indexing them locally, and ensuring relevance—is defensible against competitors adding similar functionality to their platforms.