Rasa

Rasa is a platform for product teams building customer facing AI agents in text and voice. Agents can be designed in code or in a visual builder, so engineers work in a structured, version controlled codebase while other team members build, review, evaluate and monitor visually, with changes flowing back into the shared repository. The workflow starts with a prompt that can be tuned in the interface, in code or with a coding agent, with the option to add explicit order and conditions where needed. Rasa then runs the agent through simulated conversations in voice and text, supports end to end checks and version comparison before release, and reads live conversations to tag and cluster issues. The platform lists dialogue management, a no code flow builder, language agnostic natural language understanding, enterprise search, a response rephraser, a custom actions SDK, multi LLM management and an LLM fine tuning recipe. Deployment features include Kubernetes through Helm, channel connectors, Redis concurrency, a Kafka data pipeline and OpenTelemetry observability, with single sign on, role based access control, PII management and Vault secrets management for governance. Rasa can run on the customer's own infrastructure, self managed on premises or in a private cloud, or as a managed service. Pricing has two tiers. The Developer Edition is free, with a license usable locally or in production, limited to one bot per company with up to 1,000 external conversations or 100 internal conversations per month, and community support through a forum. The Enterprise tier gives full platform access and premium support, and its price is available through sales.