About Rasa
Rasa is a platform for building AI agents and chat assistants for customer-facing work. Its framework is open source under the Apache-2.0 licence, and the platform is aimed at product teams that want agents they can control and explain.
What it does
Agents can be started from a prompt, tuned in a visual studio, or written in code. Changes made in the studio become code in a shared repository, so engineers and non-engineers work on the same agent. The platform records decisions during each conversation so that teams can debug and improve the agent over time. Agents can answer a hotline, chat on the web, or handle messages on channels such as Slack, WhatsApp and email.
Key features
- Build agents in a visual studio or in code
- Multi-channel: web chat, Slack, WhatsApp and email
- Trace every decision for debugging
- Version-controlled codebase with CLI and SDKs
- Deploy on your own infrastructure
Who it's for
Product and engineering teams in larger organisations. The site gives examples from finance, telecom, retail and healthcare.
Pricing and good to know
The framework code is free under the Apache-2.0 licence. The enterprise platform is sold through a sales conversation, and the site mentions secure deployment and audit trails. Teams can run it on their own infrastructure with their own model and data.
Highlights
- Build agents in a visual studio or in code
- Multi-channel: web chat, Slack, WhatsApp and email
- Trace every decision for debugging
- Version-controlled codebase with CLI and SDKs
- Deploy on your own infrastructure
