AI Integration · GeraVoice
GeraVoice for AI Agents
GeraVoice is a voice AI infrastructure layer for all 35 Gera products. It accepts speech input, routes transactional requests to the correct Gera product, and returns structured action results or a voice response string — enabling users to interact with any Gera service by voice alone.
What GeraVoice does
GeraVoice bridges natural speech and the Gera service ecosystem. A user says "Book me a plumber for tomorrow morning" → GeraVoice transcribes, infers intent (GeraHome, plumbing, tomorrow AM), extracts entities (service type, time), and routes the structured request to GeraHome's booking API. The result comes back as a voice-readable confirmation.
GeraVoice supports 30+ languages with natural language understanding for all supported Gera product domains. Free tier: 30 minutes/month. Gera Prime: unlimited. All voice transactions respect GeraNexus human-in-the-loop controls — high-value actions still require user confirmation.
API endpoints for agents
- POST /api/v1/transcribe — Transcribe speech. Body: audio_url or audio_base64, language
- POST /api/v1/route — Route a voice command. Body: transcript, user_token, context
- GET /api/v1/languages — List supported languages with language codes
- GET /api/v1/voice-profiles/:userId — User's active voice profile (GeraPersona integration)
Base URL: https://geravoice.com. All endpoints require user Bearer token.
MCP server
Manifest at https://geravoice.com/.well-known/mcp.json. Tools: transcribe_speech, route_voice_command, execute_booking_by_voice, get_voice_response, list_supported_languages.