Clinical documentation time cut 60% with AI-assisted SOAP notes

The Challenge

Coleman Health's clinical staff were drowning in documentation. Each patient session required a full SOAP note (Subjective, Objective, Assessment, Plan) to satisfy payer requirements and maintain compliant clinical records. With an average session length of 50 minutes, clinicians were spending 35 to 45 additional minutes post-session just on documentation — effectively turning every appointment into a 90-minute commitment.

The ripple effects were visible across the practice. Clinicians routinely arrived early and stayed late. Burnout metrics were climbing. Session capacity per day was capped by documentation backlog rather than scheduling availability. Meanwhile, the practice's EHR was a sunk-cost system with years of integration work behind it — a full replacement was not financially or operationally viable.

• 35–45 minutes per session in post-visit documentation time
• Clinician burnout correlated with after-hours documentation load
• Session capacity artificially limited by documentation bottleneck
• Payer audits requiring structured SOAP format, not free-form notes
• EHR migration not viable — required zero-disruption solution

The Solution

Intelligent Group deployed AiT Hosted Agents' clinical documentation module as a lightweight layer on top of Coleman Health's existing EHR. The agent ingests a session transcript (from ambient audio capture or structured session intake), applies a SOAP-structured generation prompt fine-tuned against Coleman's payer requirements, and produces a draft note ready for clinician review within 90 seconds of session end.

The HIPAA BAA was executed prior to any PHI touching the system. All LLM calls route through Intelligent Group's private AI gateway — no PHI is sent to a third-party model API. The draft note lives in a clinician-controlled queue: the clinician reviews, edits if needed, and approves. The approval action writes the finalized note directly to the EHR via API — no copy-paste, no dual-entry.

• Ambient session transcript captured via clinician's existing device
• SOAP-structured draft note generated in under 90 seconds
• All PHI processed through HIPAA-compliant private AI gateway
• Clinician review-and-approve workflow with inline edit capability
• Finalized note pushed to EHR via API — no duplicate data entry
• Zero EHR replacement or field hardware required

Results

In Their Words

Compliance & Security

All Coleman Health PHI is processed exclusively within Intelligent Group's private AI gateway. No protected health information is transmitted to any third-party LLM API. The HIPAA BAA was in place before any patient data touched the system. Access is role-gated: each clinician sees only their own note queue. Audit logs are maintained for every generation event and every approval action.

• HIPAA BAA executed prior to PHI processing
• Private AI gateway — PHI never leaves controlled infrastructure
• Role-based access control per clinician
• Immutable audit log for every generation and approval event
• Data residency in US-region Cloud Run

What's Next

Phase 2 extends the agent to treatment plan generation and progress note summarization. Phase 3 introduces a payer-specific validation layer that checks the draft note against payer-specific CPT code documentation requirements before the clinician review step, further reducing audit risk.

• Treatment plan drafting from session intake + prior notes
• Progress note summarization across patient history
• Payer-specific CPT validation layer pre-clinician-review
• Multi-location rollout to additional Coleman Health sites