---
title: "Getting Started with Minds | Minds"
canonical_url: "https://getminds.ai/guide/overview"
last_updated: "2026-07-12T19:18:41.387Z"
meta:
  description: "Scientifically validated digital brains that think and respond like real humans. Test, iterate, and validate with the exact audience you care about, in minutes."
  "og:description": "Scientifically validated digital brains that think and respond like real humans. Test, iterate, and validate with the exact audience you care about, in minutes."
  "og:title": "Getting Started with Minds | Minds"
  "twitter:description": "Scientifically validated digital brains that think and respond like real humans. Test, iterate, and validate with the exact audience you care about, in minutes."
  "twitter:title": "Getting Started with Minds | Minds"
---

Minds

April 5, 2026·Alexander Doudkin, CEO & Co-Founder

# **Getting Started with Minds**

Minds is a simulation platform that lets teams create digital brains of real-world personas to test, iterate, and validate ideas with specific audiences. Built on research and psychological models, these reusable personas help marketing, product, and agency teams run customer and expert panels faster.

## What is Minds?

Minds is a simulation platform for customer and expert panels. You create digital brains of real-world personas — a CMO at a mid-market SaaS, a Gen-Z student in Berlin, a specific industry expert — and ask them anything. You get side-by-side answers from a whole audience in minutes, not weeks.

## How it's different from a regular chatbot

A generic LLM gives you a plausible guess. A Mind gives you a real simulation.

Under the hood, every Mind is built by running roughly 100× deeper research across the public web than a standalone LLM query, then running that evidence through real psychological models — personality, values, motivations, buying behavior. The result is a profile that matches real human data with 80–95% accuracy, without any manual cleanup.

Practically, that means:

- **Validated, not improvised** — responses are grounded in the persona's actual worldview, sources, and psychometric profile, not a stylistic imitation.
- **Comparable at scale** — ask one question to 8, 15, or 100 Minds and get structured, aggregated answers with distribution charts.
- **Reusable** — the same panel you build today can be queried again next week with a new stimulus.

## Core concepts

### Minds

A Mind is a digital brain built from LinkedIn profiles, websites, PDFs, raw notes, or a plain description. Each one has its own knowledge base, personality, voice, and profile. You can edit everything from the context tab at any time.

### Groups

A Group is a panel of Minds queried together. Build one by picking existing Minds from your library, or describe your audience (e.g. _"enterprise CTOs in San Francisco, SaaS, 500+ employees"_) and let Minds generate a representative sample for you to iterate on.

### Chats

One unified chat is where you interact with everything. Mention a single Mind for a deep one-on-one, drop in a group for a panel, or run two groups side-by-side to see where segments agree and where they diverge. Drag in PDFs, screenshots, pitch decks, competitor creative — anything you want them to react to.

### Panels

When you query a Group, responses are auto-classified and aggregated:

- **Scale** (1–10 ratings) → distribution + group averages
- **Categorical** (yes/no, multiple choice) → percentage breakdowns
- **Qualitative** (open-ended) → clustered themes

## Who uses Minds for what

- **Marketing teams** paste in ad copy, positioning, and landing pages before they spend a dollar on media.
- **Agencies** drop in client briefs, competitor messaging, and concept work — validated insight in hours, sometimes demoed live in the pitch meeting.
- **Product teams** run feature specs, pricing tiers, and onboarding flows past target users before a single line of code. Validation cycles drop from 4–6 weeks to same-day.
- **Founders** query an expert panel directly: _should I expand to this market, is this pricing right, does this pitch land?_ Enterprise-grade advice at SMB budgets.

## Quick start

1. **Create a Mind** — click **New Mind** in the sidebar, describe the persona (add a LinkedIn URL, file, or keywords for grounding), hit **Create**. You can watch it build live in the Minds panel.
2. **Build a Group** — click **New Group**, then either drag in existing Minds or describe your target audience and review the drafted panel before it builds.
3. **Open a chat** — mention your Group in the smart input, drop in any stimulus (PDF, image, text).
4. **Ask your question** — see all Minds respond in parallel with an aggregated breakdown.
5. **Go deep** — click any Mind in the response to pull them into a one-on-one and probe further.

## Next steps

- [Minds](https://getminds.ai/guide/minds) — how Minds are built and what makes them accurate
- [Creating a Mind](https://getminds.ai/guide/my-mind) — the full creation flow, knowledge base, voice, sharing
- [Chats](https://getminds.ai/guide/chats) — the unified chat interface and deep-dive pattern
- [Panels](https://getminds.ai/guide/panels) — running multi-group research and reading the results
- [Marketplace Partner Groups](https://getminds.ai/guide/marketplace-partner-groups) — publish free public groups or sell paid Team add-ons
- [Feature Catalog](https://getminds.ai/guide/features) — feature categories and customer-facing descriptions
- [Integrations](https://getminds.ai/guide/integrations) — ChatGPT, Claude, Cursor, Google Chat, embed widget
- [MCP reference](https://getminds.ai/mcp/overview) — use Minds from any MCP-compatible AI client