Build a custom CRM for your startup with fields and workflows tailored to your exact sales process.
Deploy a self-hosted customer management system using Docker so your data stays on your own servers.
Create AI-powered tools that read and write CRM data automatically when deals or contacts change.
Version-control your CRM schema alongside your codebase so schema changes go through code review.
Requires Docker and likely a database setup; GraphQL backend needs to be running alongside React frontend.
Twenty is an open-source CRM (Customer Relationship Management) platform designed as an alternative to proprietary systems like Salesforce. A CRM is software that helps businesses track their relationships with customers, recording contacts, deals, conversations, tasks, and sales pipelines, so nothing falls through the cracks. What distinguishes Twenty from typical CRM tools is that it is built for technical teams who want to customize and extend their CRM rather than be constrained by what a vendor allows. Instead of configuring a CRM through click-based admin panels alone, Twenty lets developers define new data objects, fields, and views directly in TypeScript code, then deploy those changes to their workspace. The CRM schema is version-controlled alongside the rest of the codebase, so changes can be reviewed, reverted, and tracked just like any other software change. The core features cover the standard CRM toolkit: contacts, companies, deals, tasks, and notes. On top of that, Twenty provides workflow automation to trigger actions when records change, and built-in AI agent support so teams can build AI-powered tools that interact with their CRM data directly. Layouts and views are customizable per team. Twenty is available as a hosted cloud service (sign up and start immediately), as a self-hosted installation using Docker, or as a developer SDK where you scaffold, build, and deploy custom applications on top of the platform. A startup, SaaS company, or software team would use Twenty when they need a CRM that fits their specific data model and processes, and they want to own and extend that tooling rather than paying enterprise pricing for something rigid. The tech stack is TypeScript on both front and back ends, with a GraphQL API, and React for the user interface.
Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.