Pull data from Customer.io (Campaign Metrics, Customer Attributes, Reporting Webhooks) → analyze with Mavera (Mave Agent, Custom Personas, Chat, Generate) → get persona-validated messaging strategies, audience clusters, and automated retention research.Documentation Index
Fetch the complete documentation index at: https://docs.mavera.io/llms.txt
Use this file to discover all available pages before exploring further.
Customer.io API — Two APIs: Track (
https://track.customer.io/api/) uses HTTP Basic auth (Site ID : API Key), ~1,000 req/sec. App (https://api.customer.io/v1/) uses Bearer Token, 10 req/sec. Campaign metrics live in the App API; event tracking and customer attribute reads use Track.Prerequisites
Customer.io API credentials
Create a Track API key (Site ID + API Key) and an App API key (Bearer Token) from your Customer.io workspace settings.
Mavera API key
Get your key from Mavera dashboard.
Jobs
| # | Job | Customer.io Data | Mavera Surface | Output |
|---|---|---|---|---|
| 1 | Campaign Event Analysis → Messaging Strategy | Campaign metrics | Mave Agent + Generate | New message variations |
| 2 | Customer Attribute Personas | App API customer attributes | Custom Personas | Attribute-clustered personas |
| 3 | Webhook Event → Mave Research Trigger | Reporting webhooks | Mave Agent | Retention research |
Rate Limits & Production Notes
| Endpoint | Limit | Strategy |
|---|---|---|
| Track API (identify/track) | ~1,000 req/sec | Generous; batch where possible |
| App API (campaigns, customers) | 10 req/sec | 120ms delay; retry on 429 |
| Segment membership | 10 req/sec | Paginate with next cursor |
| Reporting webhooks | N/A (push) | Queue high-volume events |
- Store
CIO_SITE_ID,CIO_API_KEY,CIO_APP_KEY, andMAVERA_API_KEYin a secrets manager — never commit tokens. - Implement webhook signature verification with
CIO_WEBHOOK_SECRETin production. - For webhook receivers, deploy behind a load balancer and use a message queue to decouple event ingestion from Mave research calls.
- Monitor Mavera credits at Dashboard.
- Cache customer attributes locally if running persona creation frequently — App API rate limits make per-customer lookups expensive.