Detox intake requires staff to manage emotionally complex calls while also screening for eligibility, collecting information, coordinating assessments and determining when escalation is needed. For College Medical Center, the challenge was capacity: how to make sure every call was answered and moved forward, even after hours, without adding unnecessary burden to an already-busy intake team. A person calling about detox may be frightened, uncertain or acting during a brief window when they are ready to seek treatment. A family member may be trying to help without knowing what happens next. In those moments, voicemail is more than an inconvenience. It can interrupt the path to care.

Rather than replacing the intake team, GraceAI extends its capacity, coordinating the detox-intake journey across administrative touchpoints. Grace answers, screens, schedules and documents, then connects the caller with a person when human support is the right next step. Grace operates as a coordination layer across the whole detox-intake ecosystem: around-the-clock answering, eligibility screening and assessment scheduling for people seeking treatment and their families; call management, demographic and insurance capture and escalation for intake and front-office staff; symptom and assessment documentation support and urgent-call identification for clinical teams; and insurance-information capture and pre-authorization support for payers. Grace is delivered through VitalCX’s hybrid model, combining GraceAI orchestration with live healthcare guides. The deployment was deliberately scoped to a single intake line and supported by onsite implementation at College Medical Center. Screening and routing logic, escalation pathways, after-hours coverage and reporting were built and validated against real interactions before the model was expanded. Human handoffs are an intentional part of the model, particularly when a caller requests staff support or the conversation requires it.

Across the March 15 – July 2, 2026 production window, Grace handled 954 inbound intake calls from an estimated 478 unique callers, including 532 scheduling conversations. A 66.6% containment rate means roughly two-thirds of inbound calls did not require a live transfer, allowing the intake team to focus on conversations that needed a person. 280 calls were fully resolved end-to-end, 114 assessment appointments were confirmed and 99 after-hours and weekend calls were covered. Of the 319 calls routed to staff, 214 were caller-requested, demonstrating that Grace connects people to people appropriately rather than forcing automation where it does not fit. The 241 callbacks identify a clear opportunity for the next optimization cycle: moving more callers directly into scheduling could improve first-contact resolution and further reduce staff dependency. This is the first production implementation of GraceAI in a behavioral-health detox-intake setting, and it gives College Medical Center a foundation for expanding Grace into additional workflows and service lines as needs evolve.