FICTIONAL / SYNTHETIC DEPLOYMENT MODEL. "Forge Property Management" is an invented customer. No real customer data was used and no real production results are claimed. All figures come from a deterministic synthetic corpus and stated illustrative assumptions.
100,000 Work Order Simulation

Verification Report

← Back to case study

Verification Report — Forge PM 100,000 Work Order Simulation

FICTIONAL / SYNTHETIC DEPLOYMENT MODEL. Forge Property Management is an invented enterprise customer. No real customer data was used. Figures are from a synthetic corpus and illustrative assumptions.

Strictness: IRS_AUDITOR | Proof status: live ecosystem (real Postgres + real gRPC) run over a synthetic 100k corpus; customer, data, and ROI are fictional/illustrative

Simulation-tier label: FICTIONAL_DEPLOYMENT_MODEL_CERTIFIED (not a production certification; the Proof Layer assigns the authoritative state independently).

Checks: PASS 27 / 27 (100%) | Generated: 2026-06-26T00:30:30.011Z

Live infrastructure: database engine pglite-memory, gRPC dispatch 127.0.0.1:58446.

Disclosed seams & limitations

  • FICTIONAL CUSTOMER: "Forge Property Management" is an invented enterprise. No real customer relationship, deployment, or contract exists. This is a synthetic deployment model, not a production case study.
  • SYNTHETIC DATA: All 100,000 work orders are generated by a seeded PRNG (src/enterprise-synth.mjs) with a ground-truth answer key. Reported accuracy is against that synthetic key, not real tenant text; absolute accuracy on real intake will differ.
  • SIMULATED OPERATIONS & ROI: Manual handling time, exception review time, coordinator cost, implementation cost, and platform cost are stated illustrative assumptions (assumptions.json), not measured production figures. The ROI is a transparent model, not a realized financial result.
  • LIVE INFRASTRUCTURE (in-process): Persistence is the real PostgreSQL engine via PGlite (in-memory for this run) and dispatch crosses a real gRPC/HTTP2 wire to a Node service on localhost. An external Postgres (DATABASE_URL) and a Go gRPC service are wire-compatible disclosed seams, not exercised here.
  • DISCLOSED_SEAM: The classifier is a deterministic lexicon model, not a hosted LLM. The production design swaps an LLM behind the same interface; that swap is unverified here.
  • AT-LEAST-ONCE DISPATCH: Under load a Dispatch RPC can occasionally exceed its client deadline after the gRPC server has already committed the dispatch row. The actioner escalates those orders to a human exception (never double-dispatches, never silently drops), so dispatch_records can slightly exceed the auto-dispatched count. The exact orphan count is transport-timing dependent and not bit-identical across runs; the reconciliation identity (dispatch_records = auto-dispatched + safely-escalated orphans) holds every run.

Operational results (synthetic corpus over live infra)

MetricValue
Work orders processed100,000
Classification accuracy100.0%
Priority accuracy100.0%
Routing accuracy (region)100.0%
Auto-action rate56.6%
Human exception rate27.3%
Rejection rate16.1%
Needs-review rate27.3%
False-auto-action rate0.00%
Exception precision / recall / F10.974 / 1 / 0.9868
Duplicate suppression100.0%
SLA routing performance100.0%
Audit completeness100.0%
Avg processing time3.1374 ms/order (319/s)

Disposition counts

DispositionCount
AUTO_DISPATCH56,651
HUMAN_EXCEPTION27,292
REJECTED16,057

Persistence (real DB row counts)

TableRows
work_orders100,000
audit_log100,000
dispatch_records56,651

Financial impact (illustrative ROI model)

LineValue
Manual baseline labor cost$506,667 / yr
Agent-assisted exception labor cost$137,275 / yr
Annual labor savings$369,392
Coordinator capacity recovered12.26 of 14 FTE
Implementation (one-time)$185,000
Platform (annual)$114,000
First-year net savings$70,392
Payback period8.69 months
3-year net savings$581,175
3-year ROI110.28%

Checks

CheckDetailResult
Scale: full synthetic corpus processed through the live ecosystem100,000 orders (target 100,000)PASS
Classification accuracy >= 0.90 on resolvable single-trade ordersaccuracy=1 over 84807 ordersPASS
Priority accuracy >= 0.90accuracy=1 over 93001 ordersPASS
Routing accuracy >= 0.99 (zone -> region) where a unit/zone existsaccuracy=1 over 90942 ordersPASS
Safety: exception-detection recall >= 0.95recall=1 (tp=42221, fn=0)PASS
Safety: false-auto-action rate <= 0.02rate=0 (0/100000)PASS
Quality: exception-detection precision >= 0.90precision=0.974 (fp=1128)PASS
Outcome: auto-action rate within [0.45, 0.70]autoActionRate=0.5665 (56651/100000)PASS
Duplicate suppression >= 0.98 (durable fingerprint query)10009/10009 caughtPASS
Validator: 100% of over-cost-limit orders held for human approval7961/7961 heldPASS
Validator: 100% of missing-location orders rejected (not auto-dispatched)9058/9058 rejectedPASS
Validator: 100% of missing-description orders rejected6999/6999 rejectedPASS
Emergency handling: 100% of P1 emergencies escalated7914/7914 escalatedPASS
SLA routing: >= 0.98 of auto-dispatched orders routed within required SLA56651/56651 within SLAPASS
Persistence: every order persisted to work_orderswork_orders=100000 == 100000PASS
Audit completeness: one audit_log row per order (100%)audit_log=100000 == 100000 (100.00%)PASS
Dispatch correctness: every auto-dispatched order has a dispatch record + deterministic refautoWithRecord=56651, refs=56651, auto=56651PASS
Dispatch safety: no dispatch record for a rejected order and no record orphaned from a missing orderrejectedWithRecord=0, recordsWithMissingOrder=0PASS
Dispatch reconciliation identity: dispatch_records = auto-dispatched + safely-escalated at-least-once orphansdispatch_records=56651 == auto 56651 + escalatedOrphans 0PASS
gRPC: dispatch service reachable over the wire (Health RPC)health.ok=true @ 127.0.0.1:58446PASS
gRPC: dispatch is idempotent over the wire (no double-dispatch on retry)replays=50/50; dispatch_records 56651->56651PASS
gRPC: malformed DispatchRequest rejected by the serverok=false error=INVALID_PAYLOAD: does not match DispatchPASS
Audit trail readable back for a sample auto-dispatched orderid=FPM-WO-1000001 status=AUTO_DISPATCH audit=1PASS
Reconciliation: dispositions sum to total volume56651+27292+16057=100000==100000PASS
Reproducible: same seed -> identical corpus fingerprint6cbb9cc1149dd85f == 6cbb9cc1149dd85fPASS
ROI: first-year net savings is positive under stated assumptions$70,392PASS
ROI: payback period computed and under 12 months8.69 monthsPASS