Q3 2026
| Area | Item |
|---|---|
| Multimodal | Vocal Acoustic Engine v2 — extended prosodic features with sub-second windowing |
| Multimodal | Neurobehavioral Construct v2 — refinements to the 15 documented constructs with updated feature contributions |
| Agentic | CDI Agent — outpatient specialty rule packs (BH, primary care, dermatology) |
| Integrations | EHR write adapter — write-back to NextGen and eClinicalWorks |
| Language | Spanish (US) intake — validated translation pack for PHQ-9, GAD-7, AUDIT-C |
Q4 2026
| Area | Item |
|---|---|
| Compliance | HITRUST certification — final report |
| Agentic | Co-Therapy v2 — group session multimodal fusion |
| Agentic | PA Appeals v2 — auto-generated peer-to-peer talking points |
| Integrations | FHIR R5 read/write adapter |
Future
| Area | Item | |
|---|---|---|
| Multimodal | Async video pipelines — frame-rate-adaptive multimodal analysis on uploaded video | |
| Security | Speaker verification on every voice agent (voice-cloning prevention) | |
| Residency | Full EU stack with EU-only sub-processors | This roadmap is informational and subject to change. See the Change Policy for breaking-change commitments. |