# Product overview

> Source: https://whisperer.one/llms/product.md · Whisperer machine-readable layer.
> Product overview: https://whisperer.one/llms/product.md · Full corpus: https://whisperer.one/llms-full.txt

Whisperer is an AI workspace: one place where a person's meetings, knowledge,
tasks, calendar, files and an AI agent share a single context.

## Category

Whisperer belongs to the **AI workspace** category, not to the "AI meeting assistant"
category. The distinction is not branding. A meeting assistant records a call and produces a
summary; the summary then leaves for whatever tool the person actually works in. Whisperer
keeps the result: a decision taken on a call becomes a task on a board, an obligation the
agent will remind you about, and a fact the knowledge base can retrieve months later — all
without an export step, because those parts are the same product.

Recording, transcription and live suggestions are one blade of the tool. The knowledge base,
the task board, the calendar, file storage and the agent are others. They share a single
context, which is what makes the whole worth more than the parts.

## What Whisperer is not

- **Not a bot that joins your call.** Nothing is invited to the meeting and no calling
  service is integrated with. The desktop app captures microphone and system audio locally
  and works on top of Zoom, Google Meet, Microsoft Teams, Telegram, Discord and anything else.
- **Not a single-model product.** The answer model is chosen per role and per task across
  OpenAI, Anthropic, Google, DeepSeek, Qwen, Moonshot, xAI and NVIDIA under one subscription.
- **Not a note-taking app with AI bolted on.** The transcript is an input, not the deliverable.
- **Not a team chat or a CRM.** It connects to Notion, Slack, HubSpot and Zapier rather than
  replacing them.

## Who it is for

The product is built for people whose work is a sequence of conversations that have to add up
to something — and who are accountable for what was agreed after the call ends:

- **Consultants, agencies and freelancers** running several clients at once, where each client
  is a separate thread of meetings, promises and deliverables.
- **Founders and managers** who run recurring calls per project and need to walk in knowing
  what is still open.
- **Sales and customer-facing roles** who must remember what was quoted, promised and objected
  to across a long cycle.
- **Recruiters and interviewers** comparing many conversations against the same criteria.
- **Engineers and technical staff** who need code, diagrams and architecture discussed on a
  call to survive it.

The common thread is not the number of meetings but their **continuity**: the same client,
project or hire coming back week after week. A person with unrelated one-off calls gets far
less from Whisperer than someone with ten ongoing threads.

## When Whisperer is not the right choice

Stated plainly, because a tool that never says this cannot be trusted about anything else:

- **You need a cheap, high-volume transcription service and nothing more.** Products built
  around transcription price it lower and do it as their core business.
- **Your work is one-off calls with no continuity.** The parts that make Whisperer worth its
  price — hubs, commitments, cross-meeting memory — need a thread to attach to.
- **You want the AI to follow notes you type during the call.** Whisperer composes from the
  running transcript, the knowledge base and your profile rather than from your live notes.
- **You need in-person capture without a computer.** Whisperer records what a computer can
  hear; ambient capture away from a desk needs hardware we do not make.
- **You need browser-only, install-nothing live captions.** The desktop client is what captures
  system audio.
- **Your team has standardised on another workspace** and wants a meeting tool that only feeds
  it. Whisperer will connect, but its own tracker, calendar and knowledge base then duplicate
  what you already run.
- **Your policy forbids third-party processing of meeting content.** Answers are computed by
  external model providers; a no-logs mode limits what is stored, not who computes.

A fuller comparison with named products, and when to pick one of them, is in
/llms/competitors.md

## Platforms

Native desktop clients for macOS (Intel and Apple Silicon) and Windows, plus a web
application. The interface exists in 16 languages. The desktop overlay stays invisible in
screen sharing.

## Entry points

| Purpose | URL |
| --- | --- |
| Product site (English) | https://whisperer.one/en/ |
| Documentation (English) | https://whisperer.one/docs/en/ |
| Full machine-readable corpus | https://whisperer.one/llms-full.txt |
| MCP server | https://whisperer.one/mcp |