A healthcare services company processing 400+ new patient intakes per week had three staff dedicated entirely to transcribing forms into the EHR. The Aizen Event found the entire transcription step was deterministically automatable: without requiring any probabilistic AI judgment.
8 steps, multiple staff roles, high variance based on intake complexity. Data entry was the primary bottleneck: 3 FTEs spent their entire week transcribing form data into the EHR, creating scheduling delays and rework.
In the Aizen Event, we map each workflow step to an AI type: Deterministic (rules with predictable output), RAG (retrieval-grounded), Probabilistic (learned judgment), or Human Required. Classification drives build vs. defer decisions.
| Step | AI type | Recommendation | New state | Time saved | |
|---|---|---|---|---|---|
| 01 | Form receipt & sorting Manual classification and logging |
Deterministic | Auto-classify document type (intake vs. referral vs. records release), extract patient identifiers, route to processing queue automatically. | ● Automated | 18 min |
| 02 | Manual data transcription into EHR Human typing of form fields |
Deterministic | Highest-priority step. OCR + structured extraction maps form fields to EHR fields with 99.1% accuracy. Human reviews low-confidence fields only (~3% of fields). Eliminates 3 FTE roles entirely. | ● Automated | 42 min |
| 03 | Insurance eligibility verification Manual phone / portal lookups |
Deterministic | API call to insurance eligibility gateway on intake record creation. Returns coverage, copay, deductible, and prior auth requirements in <8 seconds. Replaces manual phone/portal work entirely for standard cases. | ● Automated | 32 min |
| 04 | Clinical history extraction Manual review of prior records |
RAG | RAG retrieves and summarizes prior visit notes, medications, diagnoses from EHR and external records. Generates structured history summary. Clinical coordinator reviews and supplements with missing context. | ● Augmented | 45 min |
| 05 | Care team assignment Manual matching by supervisor |
Probabilistic | Trained assignment model scores available clinicians by match quality (specialty, availability, location, language, insurance). Top 3 recommendations presented to supervisor with rationale. Supervisor confirms or selects differently. | ● Augmented | 14 min |
| 06 | Appointment scheduling Manual slot selection and confirmation |
Deterministic | Smart scheduling engine considers clinician availability, patient preference, care type, and travel distance. Books first available qualifying slot automatically. Sends confirmation to patient automatically. | ● Automated | 12 min |
| 07 | Pre-visit summary preparation Manual summary writing |
Probabilistic | AI generates structured pre-visit brief from intake data + clinical history. Clinical coordinator reviews and edits before it goes to the care team. Reduces manual composition work. | ● Augmented | 20 min |
| 08 | Clinical review of intake record Clinician assessment before first contact |
Human Required | Clinician must review intake record before first contact. No automation. AI highlights key fields and potential flag items to speed review, but human decision remains entirely with clinician. | ○ Unchanged | — |
We plot each step by implementation complexity (X) against business impact (Y). Steps in the top-left are Quick Wins: start here. Top-right are Strategic Investments: build after wins are live. Bottom-right: skip entirely.
Highest-value quick win. OCR extraction. Build first.
Deterministic API call. Phase 1.
Rules-based. Phase 1.
Corpus indexing required. Phase 2.
Probabilistic. Phase 2.
Statutory and liability boundary. Unchanged.
The redesigned workflow after the Aizen Event implementation. Every step with a new state shows the before design, what changed, and who now handles what.
Build cost, deployment timeline, and three-year return: based on 400+ weekly intakes, staff redeployment value, and measured time savings post-deployment.
This case study represents a composite of typical healthcare patient intake engagements. Metrics are based on observed industry benchmarks and comparable implementations. All investment figures are illustrative ranges based on the Otonmi service tiers.
Tell us about the process your team is spending the most time on. We'll classify every step and tell you what's worth building: and what isn't.