Dograh Docker setup and comparison with Vapi

Dograh is presented as an open source voice AI platform with a visual workflow builder and support for self-hosting. The tutorial starts with a local Docker setup: clone the repository, enter its directory, and run Docker Compose up. Running the platform locally does not establish that its speech, model, or phone services work offline; the speaker describes connecting providers for those parts.

The example is a lead qualification AI agent for inbound calls. Its flow asks about the caller's project, company size, and budget, then adds an API tool call, a branch, and a transfer. The speaker proposes using the API step to create or update a CRM lead and transferring qualified callers to a human. After a test call, the walkthrough examines the transcript, trace, tool activity, state changes, and recording.

The comparison focuses on infrastructure control. The speaker describes Vapi, Bland, and Retell as convenient hosted options, with tradeoffs around pricing changes, limits, and custom deployment. Pipecat and Vocode appear as frameworks that allow more control but require developers to build the surrounding interface and workflow tooling. Dograh sits between those approaches in the speaker's assessment: a visual builder with provider choice and inspectable runtime behavior. The speaker also calls the project new, so the demonstration should not be read as evidence of production maturity.