Case Study 02 — Multi-Agent Care Operations Platform

Intelligent care
coordination,
end to end.

Designing a multi-agent AI platform that guides clinic staff through every patient touchpoint — from scheduling and billing to post-operative follow-up and insurance appeals.

Role
UX Lead / AI Experience
Industry
Healthcare — Ambulatory Care
Team
2 Designers · 4 Engineers
Timeline
6 months · 2024–2025
The Problem
Clinic staff were drowning in
manual coordination work.

Every patient interaction required staff to manually pull records, verify insurance, coordinate between providers, and document outcomes across disconnected systems. High cognitive load led to errors, delays, and patient dissatisfaction — especially at surgical handoff points.

01
Fragmented scheduling workflows
Front desk staff had no live guidance during patient calls — every decision relied on institutional memory and manual lookup.
02
Insurance verification bottlenecks
Billing teams spent hours manually verifying coverage and calculating estimates before each surgical procedure.
03
Missed post-operative follow-up
Clinical staff had no systematic way to reach patients after surgery — follow-up appointments were often missed or delayed.
04
Insurance denials with no structured response
When claims were denied, staff had to manually draft appeal letters — time-consuming, inconsistent, and often unsuccessful.

Five roles. One continuous loop of care.

The patient sits at the center of every decision. Four clinical roles coordinate around them, with AI agents handing off context at every step — forming a closed loop that always returns value back to the patient.

THE PATIENT
At the center
of every decision
THE CORE
At the center of care
Every handoff, every decision, every automation is designed to return value to the patient — with transparency, safety, and dignity at the core.
Front Desk
Scheduler
Capture layer
Speech-to-Text (STT) · Live Transcript
Agents
TipsAgentScheduleAgentDoctorNotesAgent
01 — ENTRY
Patient calls to schedule surgery
Verifies identity, confirms physician, and books the appointment — guided by real-time coaching that surfaces the right action at the right moment. Every manual lookup becomes an automated handoff.
Billing
Financial Coordinator
Capture layer
Speech-to-Text (STT) · Live Transcript
Agents
TipsAgentBillingAgentEstimateCostAgent
02 — PRE-SURGERY
Insurance verification & cost estimate
Auto-triggered after scheduling. Verifies coverage, calculates out-of-pocket costs, and sends the patient a clear financial summary before surgery day — eliminating pre-surgical billing surprises.
Clinical
Nurse / Care Manager
Capture layer
Speech-to-Text (STT) · Live Transcript
Agents
TipsAgentClinicalAgentScheduleAgent
03 — POST-OP
Recovery check-in & follow-up
Reaches out after surgery, assesses recovery within clinical guardrails, and autonomously books follow-up appointments based on protocol — ensuring no patient falls through the cracks.
Insurance
Appeals Coordinator
Capture layer
No STT · Document-driven
Agents
ClaimAgentAppealAgent
04 — RESOLUTION
Claim denial → AI-drafted appeal
When insurance denies a claim, drafts a structured, evidence-backed response from clinical notes and policy data. Staff review and send, rather than write from scratch.
↻ Hover over any role to reveal the story
01 — ENTRY
Front Desk
Scheduler
Capture layer
Speech-to-Text (STT) · Live Transcript
Agents
TipsAgentScheduleAgentDoctorNotesAgent
Tap to reveal story ↻
01 — ENTRY
Patient calls to schedule surgery
Verifies identity, confirms physician, and books the appointment — guided by real-time coaching that surfaces the right action at the right moment. Every manual lookup becomes an automated handoff.
THE PATIENT
At the center of every decision
← Swipe to cycle through roles →
Live demo — real Nova-3 Medical STT, Aura-2 TTS, multi-agent care call.

Two end-to-end calls powered by Deepgram. Aura-2 plays the dialogue while Nova-3 Medical transcripts drive the right-panel checklist; HITL popups pause audio at every decision point so you drive the agents yourself.

Scene 01·Surgery Scheduling Call
Standby · Nova-3 Medical
Conversation Transcript
Calling patient...
Live Coaching
Real-time action prompts
1Patient connectedTips
2Verify identityTips
3Confirm surgery & providerTips
4Schedule patient appointmentSchedule Agent
5Send confirmation emailEmailAgent
6Setup reminderTips
7Route note to providerHealthAgent
8Schedule billing consultationSchedule Agent
Powered by
Aura-2 TTSNova-3 Medical STT
Multi-Agent Architecture
Four agents. One orchestrated care workflow.

Each persona runs on its own bundle of agents inside the CareAssist platform. Click any use case to see the screen that persona works in.

Patient
Calls clinic
Resolved
Care complete
Live transcript → real-time coaching
Speech-to-text captures the call as it happens; TipsAgent listens and surfaces the next action for staff. Used in Scheduler, Billing, and Nurse workflows — Appeals is fully document-driven.
Outcome & Impact
60%
Reduction in scheduling call time
Real-time coaching eliminated manual lookup during calls, cutting average handle time significantly.
4x
Faster insurance verification
BillingAssist automated coverage checks that previously required multiple phone calls to payers.
85%
Post-op follow-up completion rate
ClinicalAssist ensured no patient missed a follow-up — up from an estimated 40% with manual outreach.
Designing AI that assists — not replaces — clinical judgment.

The core design challenge was establishing appropriate AI boundaries. Each agent needed clear escalation paths — moments where the system explicitly deferred to human expertise rather than proceeding autonomously. This was especially critical for clinical and billing decisions, where errors carry real consequences for patients.

The real-time coaching layer became the key design mechanism for building staff trust — giving humans visibility into what the agent was doing and why, at every step.