What it does
Real-time APIs that expose queue-level representative availability and estimated wait times. Designed for use by IVR systems, bots, and digital assistants to make routing decisions or give customers accurate queue information before connecting.
Key facts
- GA'd April 2026; enabled automatically for admins/developers: no separate toggle
- Two APIs: one for scenarios where a conversation already exists, one for pre-conversation (no existing context)
- Uses the same routing path logic as actual conversations: so the availability data reflects the real workstream/queue routing, not a generic count
- Returns: representative availability (by queue), average wait times
- Intended consumers: Copilot Studio agents, IVR flows, external orchestration systems
When to use / skip
Use this if you're building a self-service IVR or bot that needs to tell callers how long they'll wait, or if you want to route to an alternative queue based on real-time availability. Also useful for proactive SMS flows that should only connect if agents are available.
Skip if your routing is simple and wait time estimation isn't a customer experience requirement, the API adds integration complexity that isn't justified for low-sophistication deployments.
Configuration decisions
- Which API variant to call: conversation-context API (when you have a work item ID) vs. pre-conversation API (when routing hasn't started yet)
- What availability thresholds trigger different IVR paths (e.g. if wait > 10 mins, offer callback)
- Authentication: Dataverse API credentials required: standard OAuth 2.0 client credentials or delegated access
Gotchas
- "Same routing path as the conversation" means the API respects your workstream, queue assignment rules, and operating hours. If you call it outside operating hours, the returned availability reflects 0 available reps: don't interpret that as the API being broken.
- Average wait time is an estimate, not a guarantee. If you surface it to customers ("your wait time is 3 minutes"), manage expectations: actual wait may vary, especially during sudden volume spikes.
- The API is new as of Apr 2026. There's no documented rate limit at launch: worth testing under load before putting it in front of a high-volume IVR.
Consultant notes
- This is a powerful capability for IVR self-service, but it's brand new as of April 2026: no documented rate limits at launch. Test under realistic load before putting it in front of a high-volume IVR. An undocumented rate limit discovered at go-live will break your IVR's availability messaging.
- The operating hours behaviour is a simple trap: the API returns zero available reps outside business hours, not an error. If your IVR logic branches on "no available reps," make sure it handles the after-hours case gracefully rather than routing everyone to the overflow queue at 8am on Saturday.
- If you surface estimated wait times to customers in the IVR, word it carefully: "approximately X minutes" or "currently a short wait." Exact figures create unrealistic expectations. Volume spikes between the API call and the call connecting can double a wait time in 30 seconds.
Worth revisiting once Copilot Studio agent integration examples are published showing this API in action.