A

AfriRoute

Developer Portal

public api
Sign in
AI Routing

Show developers how AfriRoute chooses the best route and how to inspect the result

AI routing should be documented as part of the delivery workflow: send a request, see how route quality affects delivery, inspect routing visibility, and use dashboards or webhooks to understand what happened next.

Why developers care

AI routing is visible when delivery quality improves and the decision can be inspected

The docs should explain AI routing in the same place developers already care about sends, OTP delivery, retries, and webhook outcomes. The goal is not to describe private models. The goal is to show what signals matter, which public surfaces expose those signals, and how teams should use them in production.

AI routing uses signals like:
delivery_probability
expected_latency_ms
route_health
fallback_readiness
fraud_aware_policy

Developers should inspect the result through:
routing decisions
routing insights
delivery webhooks
operator performance pages

Signals developers should understand

Delivery success by market, route, and carrier
Latency by message type and time window
Carrier degradation and outage events
Route cost, fallback outcomes, and retry effectiveness
Spam filtering, number abuse, and OTP fraud patterns
Country and carrier-specific policy constraints

External surfaces that show routing outcomes

SMS send result

POST /api/v1/sms/send

Use the external send contract as the entry point. Routing happens behind the API boundary and returns developer-safe delivery state.

Message status

GET /api/v1/messages/{messageId}

Expose delivery state through a public message status lookup rather than private routing or operator dashboards.

OTP with priority routing

afriroute.verify.start

Verification flows can benefit from routing quality while the public contract stays focused on managed challenge status and approval.

Rollout sequence

1

Document the AI-routing story through supported sends, dashboard visibility, and delivery outcomes developers can actually inspect.

2

Connect routing decisions to real workflows: messaging, OTP verification, retries, and webhook-driven delivery monitoring.

3

Keep private carrier-control services and unpublished scoring APIs outside the public docs contract.

4

Use network and performance pages to reinforce why the routing story is credible across markets and operators.