Skip to content
AI Agents / Med-Only Workers' Comp

Med-only workers' comp claims, automated end to end.

Med-only claims are the highest-volume, most rules-bound files in workers' comp — and the ones examiners should touch least. The Layerup med-only agent runs each claim in one continuous long-horizon run: FROI intake, compensability, bill and record processing, treatment monitoring, and autonomous closure. It escalates only when the file stops being med-only — and hands off a decision-ready packet when it does.

HorizonDays to weeks
ScopeFROI to closure
AutomationFull, with escalation gates
CoverageAll US jurisdictions
01One-shot vs long-horizon

One-shot agents answer. Long-horizon agents finish.

Most AI agents are one-shot: one prompt, one response, no memory, no obligation to finish. Layerup agents are long-horizon — each owns a workflow end to end and stays on it, uninterrupted, for hours or days until the work is done.

Span
One-shot

One prompt in, one output out. The run is over in seconds — whatever state the work is in.

Long-horizon

One workflow worked continuously, for hours or days, until it is actually finished.

State
One-shot

Stateless per call. Everything learned about the case is forgotten between steps.

Long-horizon

Persistent working memory of the whole case — every document, finding, and decision carried forward.

Missing evidence
One-shot

Guesses or stalls when an input is incomplete. Someone has to notice and re-prompt.

Long-horizon

Requests the missing document, follows up, waits for it to land, and resumes exactly where it left off.

Systems
One-shot

Returns a response. Acting on it — updating the core system, moving the queue — is still human work.

Long-horizon

Reads from and writes back to your systems of record throughout the run, every action logged for audit.

Output
One-shot

A label, a draft, a score — an input to someone else's work.

Long-horizon

Finished, decision-ready work with rationale and citations attached, staged for human approval.

In this workflow

A one-shot agent can summarize a medical bill. It can't own a med-only claim: qualify it at intake, decide compensability, process each bill and record as it arrives over the following weeks, watch treatment against guidelines, detect the full-duty release, and close the file — while standing ready to escalate the moment lost time, an attorney, or a fraud signal appears. The long-horizon med-only agent carries the claim from FROI to closure in a single run.

02Lifecycle

FROI to closure — one agent, one continuous run.

Every stage below belongs to the same agent's uninterrupted run. On clean med-only files, no examiner touch is required at any stage; the examiner's queue holds only the files that genuinely need judgment.

Med-only claim lifecycle
6 stages
  1. 01
    FROI intake & med-only triage
  2. 02
    Coverage & compensability verification
  3. 03
    Medical bill & record processing
  4. 04
    Treatment monitoring
  5. 05
    MMI / release detection
  6. 06
    Autonomous closure & write-back
03Capabilities

Six stages. Zero touches on clean files.

From intake qualification to closure write-back, the med-only agent executes each stage autonomously — reasoning over the entire file at every new document, not just the latest page.

01 — Intake & med-only qualification
5 steps

Every FROI triaged, the med-only path confirmed.

The agent intakes First Reports of Injury from any channel, validates required fields, matches employer and policy, and qualifies the claim as med-only — no lost time, no indemnity exposure — before setting up the file and its jurisdiction-specific action calendar.

  • 01Multi-channel FROI intake
  • 02Med-only qualification against lost-time criteria
  • 03Employer and policy match
  • 04Jurisdiction and statutory deadline setup
  • 05Claim system file creation
02 — Coverage & compensability verification
5 steps

Compensability decided on evidence, not backlog order.

The agent verifies policy effective dates, class codes, and coverage applicability, reconciles the injury account across the FROI, employer statement, and initial medical documentation, and issues the compensability determination on clean files — routing only disputed ones for review.

  • 01Effective date and class code verification
  • 02Injury account consistency check
  • 03AOE/COE evidence assembly
  • 04Autonomous determination on clean files
  • 05Disputed-file routing with rationale
03 — Medical bill & record processing
5 steps

Bills and records processed as they arrive.

The agent ingests provider bills and medical records continuously, extracts service codes, checks fee schedules and treatment guidelines, stages adjudication, and keeps a running clinical timeline — so the file is always current without an examiner touching it.

  • 01Provider bill ingestion and classification
  • 02Service code extraction
  • 03Fee schedule and treatment guideline checks
  • 04Adjudication staging and payment prep
  • 05Clinical timeline maintenance
04 — Treatment monitoring over the claim's life
5 steps

The agent stays on the file between touchpoints.

Med-only claims develop quietly over weeks. The agent monitors treatment progression, provider utilization, and work status on every new document — reasoning over the whole file each time, not just the newest page — and flags drift from expected recovery.

  • 01Continuous treatment progression monitoring
  • 02Utilization and duration guardrails
  • 03Work status extraction on every document
  • 04Recovery drift detection
  • 05Provider follow-up prompts
05 — MMI / release detection & closure
5 steps

Closed when the medicine says so, not when the queue does.

The agent detects maximum medical improvement and full-duty release signals in provider documentation, confirms no open bills or pending actions remain, and closes the claim autonomously with a complete closure record.

  • 01MMI and full-duty release detection
  • 02Open bill and pending action sweep
  • 03Statutory closure requirement checks
  • 04Autonomous closure execution
  • 05Reopening trigger monitoring
06 — Write-back & audit trail
5 steps

Every action in your systems, with its reasoning attached.

The agent writes every determination, payment staging, note, and closure back into your claims core, with a reasoning trace on each action — so auditors, regulators, and examiners can reconstruct the whole claim without a single manual log.

  • 01Claims core write-back
  • 02Reasoning trace on every action
  • 03Hash-chained audit logging
  • 04Jurisdictional reporting support
  • 05Full file reconstruction on demand
04Escalations

Full automation — until the file stops being med-only.

Autonomy is only safe with sharp exits. The agent continuously tests every claim against the conditions below, and the moment one trips, it stops the straight-through path and hands the examiner a decision-ready packet — evidence, timeline, and rationale included.

01

Lost-time conversion

The moment medical documentation shows time away from work or restrictions the employer can't accommodate, the agent converts the file and hands the examiner an indemnity-ready package.

  • Off-work and restriction detection
  • Indemnity exposure summary
  • Wage data staging
  • Examiner-ready conversion packet
02

Attorney representation

When a representation letter or litigation signal appears, the agent freezes autonomous decisions and escalates with the full file history and a litigation-ready timeline.

  • Representation letter detection
  • Autonomous decision freeze
  • Timeline and evidence assembly
  • Counsel-ready packet
03

Treatment beyond guidelines

Treatment that exceeds duration, utilization, or medical-necessity guidelines pauses straight-through processing and routes to the examiner with the clinical rationale laid out.

  • Guideline deviation detection
  • Utilization review preparation
  • Clinical rationale summary
  • Nurse case management handoff
04

Disputed compensability

Inconsistent injury accounts, coverage gaps, or late reporting stop the autonomous path — the agent escalates with the evidence conflict laid out side by side.

  • Statement inconsistency detection
  • Coverage gap flags
  • Late-reporting analysis
  • Evidence conflict summary
05

Fraud & subrogation signals

Provider patterns, prior-history conflicts, or third-party liability signals trigger an escalation with an SIU referral or recovery packet already prepared.

  • Provider pattern detection
  • Prior history reconciliation
  • Third-party liability signals
  • SIU / recovery packet preparation
06

Statutory deadline risk

If any jurisdiction-specific deadline comes under pressure — acceptance, payment, or reporting — the agent escalates before it slips, with the required action staged.

  • Jurisdictional deadline tracking
  • At-risk action alerts
  • Required filing preparation
  • Compliance evidence trail
05KPIs

The metrics comp claims leaders manage.

Layerup is measured on straight-through closure, cycle time, cost per claim, and escalation precision — not on tokens, prompts, or messages.

Straight-through closure rate

01

Lift the share of med-only claims closed with zero examiner touches.

Claim cycle time

02

Compress FROI-to-closure from weeks to the pace of recovery.

Cost per claim

03

Remove manual handling from the highest-volume claim type.

Medical bill turnaround

04

Adjudication-ready bills within fee schedule and guidelines.

Examiner touches per claim

05

Reserve examiner time for files that need judgment.

Escalation precision

06

Escalate the right files — no missed conversions, no noise.

Med-to-indemnity catch rate

07

Detect lost-time conversion the day it appears in the record.

Statutory SLA attainment

08

Meet every jurisdictional acceptance, payment, and filing deadline.

Leakage reduction

09

Catch duplicate bills, out-of-guideline treatment, and missed recoveries.

Get started

Take med-only claims off the examiner's desk.

Deploy the med-only agent on one jurisdiction or one employer book. Prove straight-through closure and cycle-time lift, then expand to lost-time claims.