AI Provider Integrations
Connect model providers without losing channel, webhook, or governance context
Provider integration docs should show how developers configure models, send completion requests, and keep AI behavior aligned with messaging, bots, and support workflows.
Docs mode
Reference and guide
Keep the task clear, the example nearby, and the next step obvious.
Provider model
Treat providers as configurable infrastructure inside the developer workflow
The docs should make it clear that model providers are not a separate product island. They sit beside channels, bots, webhooks, and support automation so teams can keep one integration path while changing providers, prompts, or routing rules.
Provider endpoints
GET /api/v1/integrations/ai/providers
POST /api/v1/integrations/ai/providers
POST /api/v1/integrations/ai/services/completion
What developers should use this for
Generate bot responses for WhatsApp or SMS flows.
Run completion or classification inside support workflows.
Keep AI configuration close to channel and webhook behavior.
Swap or compare providers without rewriting channel integration code.
Completion example
Provider-backed completion request
POST /api/v1/integrations/ai/services/completion
{
"provider": "openai",
"model": "gpt-4.1-mini",
"prompt": "Summarize this delivery incident for the operations team",
"context": {
"channel": "sms",
"market": "ET"
}
}