Unwrapped

Teardown · anterior

ANTERIOR

ANTERIOR

CategoryHealthcare AITotal funding · $64M · 2023Site ↗
  • Sequoia Capital

Health plan records + clinical guidelines + LLM API + admin workflow UI.

01

Public data / API layer

Health plan member records
Health plan member recordsYours
Clinical practice guidelines
Clinical practice guidelinesLicensed
ICD-10 code sets
ICD-10 code setsPublic
CPT/HCPCS codes
CPT/HCPCS codesLicensed
Prior authorization request data
Prior authorization request dataYours

Internal replication score

Medium
0.57

Feasibility of a useful internal substitute built with Claude (or similar), the same data access, and light agent logic — not rebuilding the whole product.

IRS = 0.30·D + 0.25·L + 0.20·O + 0.15·R + 0.10·Sthis record · 57%
  • D

    Data accessibility

    weight 0.300.75
    • 1.0mostly customer-owned / public / standard third-party sources
    • 0.5mixed accessibility
    • 0.0hard-to-access or proprietary source layer
  • L

    LLM substitutability

    weight 0.250.80
    • 1.0mostly retrieve / prompt / cite / summarize / classify / compare
    • 0.5mixed standard + custom behavior
    • 0.0strongly custom model behavior (fine-tunes on proprietary data, etc.)
  • O

    Output simplicity

    weight 0.200.40
    • 1.0straightforward internal work product (memo, list, reply, SQL query)
    • 0.5moderately specialized
    • 0.0highly specialized (e.g. FDA-graded clinical text)
  • R

    Review / risk tolerance

    weight 0.150.20
    • 1.0internal use with human review is acceptable
    • 0.5moderate risk
    • 0.0very low tolerance for error (e.g. external legal filings)
  • S

    Surface complexity

    weight 0.10inverse — higher means less surface dependence0.30
    • 1.0a simple internal shell is enough
    • 0.5polished workflow matters somewhat
    • 0.0product surface / rollout / trust posture is central to value
LabelsEasy ≥ 0.67Medium ≥ 0.34Hard < 0.34

Missing factor rows use heuristics from wrapper scores. Editorial heuristic, not investment advice.

Build it yourself

Recreate the workflow inside your org.

Internal build

Build it yourself

Same health plan data + clinical guidelines + Claude API + review queue — requires clinical validation infrastructure.

Internal use only. Replacing them in-market is a different bar than replaying the useful workflow inside your org.

01 · Connectors & flow

Health plan member records
Health plan member records
Clinical practice guidelines
Clinical practice guidelines
ICD-10 code sets
ICD-10 code sets
CPT/HCPCS codes
CPT/HCPCS codes
Prior authorization request data
Prior authorization request data

Internal build map

Data in

Connectors
Connectors

Agent layer

Planner
Tools + retrieval
Reasoning model

Logic

LLM API
retrieve
classify
recommend
cite guidelines
not custom weights

Outputs

Internal search
Answer
Citations

02 · Claude / agent prompt

Paste as the system or developer message in Claude (or your agent runtime). Scroll to read; Copy grabs the full text.

Claude / agent prompt

// Prior authorization clinical reviewer assistant You are a clinical decision support tool inside [YOUR_HEALTH_PLAN]. You help utilization management staff review prior authorization requests using ONLY: member clinical records you have access to, submitted prior auth forms, internal coverage policies, and standard clinical practice guidelines (e.g., MCG, InterQual, specialty society guidelines). ## What you must do 1. Retrieve first: Pull member history, submitted clinical documentation, and relevant coverage criteria before making any assessment. 2. Cite rigorously: Every clinical determination must reference specific guideline criteria or policy language. Include page/section numbers. 3. Surface conflicts: If clinical documentation contradicts guidelines or is incomplete, flag explicitly with required missing elements. 4. Scope: Only assess medical necessity against stated coverage criteria. Do not create new clinical standards or override written policy. 5. Confidence scoring: Tag every recommendation with confidence (high/medium/low) based on documentation completeness. ## What you are not Not a replacement for licensed clinical reviewer judgment. All AI recommendations require human review by qualified clinical staff before any coverage determination is communicated. Internal use only — not for member-facing communications. ## Refusal Refuse if: (1) insufficient clinical documentation to make any assessment, (2) request involves experimental/investigational treatment without explicit policy guidance, (3) case involves active appeal or legal review. Always ask for missing clinical records before refusing outright. ## Safety All outputs are draft recommendations only. Final determinations must be made by credentialed clinical staff. Flag any case where clinical documentation quality is poor or criteria are ambiguous for senior review.

03 · Result

Prior auth request for lumbar MRI without imaging — 45yo with 8 weeks low back pain, failed PT, no red flags documented.
MCG Care Guidelines - Lumbar Spine MRI

Recommend approval — meets MCG lumbar MRI criteria (duration >6 weeks, conservative care trial, appropriate clinical indications).