Palette
Palette is an AI-powered user research platform that helps product teams design, launch, and analyze studies directly within their product. It supports user studies, usability tests, and microstudies, with AI agents that can build studies automatically and moderate interviews. The platform also offers an MCP server that lets AI coding agents query research data and launch studies from tools like Cursor and VS Code.
- Data & Analytics
- AI Agents
- Artificial Intelligence
- Enterprise Software
- Software Only
Funding
Founders
Founder details are not available yet.
Product
Problem
Product teams often struggle to gather and synthesize user feedback consistently across interviews, support tickets, and product analytics. Insights get scattered across tools, lose context, and lack the structured evidence needed to confidently prioritize which features to build next.
Solution
Palette provides an AI-powered research platform that helps teams gather and analyze user insights to make better product decisions. The platform supports multiple study types, including user studies for exploring what people think and need, usability tests for seeing how people complete tasks, and microstudies for asking quick questions while users are in the product. Palette's agentic design allows teams to create studies manually or let AI agents build them automatically, while recruitment can draw from a company's own users or Palette's external participant panel. AI moderation enables follow-up questions during interviews to surface deeper insights, and the platform automatically synthesizes responses into structured insights that can be reviewed through charts, tables, and linked summaries.
Target Audience
Primary users are product managers, UX researchers, and engineering teams at software companies who need continuous, evidence-backed user insight to prioritize and validate feature development.
Features
- AI-moderated user studies that recruit real participants and ask follow-up questions during interviews to surface deeper insights
- Usability testing with per-task success rates, median completion times, and ranked friction issues citing the affected participants
- Inline microstudies served via a JavaScript snippet that recruit and segment visitors on the company's own site without leaving the page
- Question types include multi-select, written response, single select, scale, ranking, and recorded response
- In-app invites and page-based targeting to choose where study invitations appear, from user settings to checkout
- Custom participant groups for inviting only specific user segments, such as active users, subscribers, or new users
- Rewards system that automatically grants credits or benefits in the product for qualifying completed responses
- Palette MCP Server connecting AI coding agents in Cursor, VS Code, or Claude Code to workspace research with granular Read, Write, and Commission permission keys
- Heatmap feature in development to show where users click and how they move through a page