Built by Emcie for teams deploying reliable customer-facing agents in regulated industries.

Uses condition-action guidelines, journeys, and contextual filtering instead of large prompts to control agent behavior.

Used as a Python framework embedded into customer support, sales, and onboarding applications.

Open sourced around 2024 and actively maintained with production deployments reported through 2026.

Keeps conversational agents explainable and compliant for finance, healthcare, and insurance use cases.

Depends on an underlying LLM provider and integrates tools the agent can call during conversations.

Open source under Apache 2.0 license, runs locally or self-hosted, developer controls where conversation data is sent.