AVA Outbound - New Voice Model
Following a few weeks of beta testing, the new Realtime model was applied to all outbound calls for all AVA practices. The new voice has several benefits over the previous voice:
- Decreased call latency at call beginning (~5s decrease)
- Higher callback rate, more texts being sent
- Higher voicemail detection accuracy and voicemail reporting
- Higher appointment booking rate (2.5x)
SMS Enhancements
A new shortcode was enabled for all VetPawer accounts. The shortcode has higher delivery rates and lower spam detection rates compared to the Toll-Free Number. The use of the shortcode also provides us with more accurate delivery tracking, to see if a message was successfully delivered or failed to reach the client.
AVA Scheduler - GCP Performance Improvements
Deployed performance improvements to the Google Cloud Platform configuration to prevent chat widgets or AVA scheduling configurations failing to load.
AVA - Provider Selection Refinement
Users can now specify a doctor they don't want. Additional verifications to ensure the LLM's selected provider matches the provider associated with the opening.
AVA Dashboard - Reporting Updates
A delay in generating AVA metrics was detected, which resulted in the dashboard not updating in a timely manner. Enhancements to the generation process were deployed to remove this delay.
AVImark DataLink
A correction to the AVImark DataLink integration was deployed for blockoff exceptions. If blockoff exceptions were managed by room, the room was not properly applied so the exception was read and applied to all exam rooms, leading to appointment times being inadvertently available. A new version of the DataLink was deployed to address this.
Comments
0 comments
Please sign in to leave a comment.