LINOSCORE
AI ERP Applications
Meter-Read Reconciliation · Anomaly Detection

Clarity Utility

Month-end utility processing without the spreadsheets.

Third-party water, gas, and electric bills ingested, meter reads reconciled, and usage anomalies flagged against your 12-month rolling average. Built to process a full municipal or campus month in minutes instead of days.

97%
extraction accuracy
±5%
default anomaly tolerance
batch volume, async
Capabilities

What Clarity Utility does.

Multi-provider intake

One pipeline across your water, gas, electric, and telecom providers. Each provider has its own schema variant; the core platform is the same.

Meter-read extraction

Meter start, meter end, usage period, units (CCF, kWh, therms), and any demand-based or tiered charges. Units normalised for reporting.

Anomaly detection

Usage compared against a 12-month rolling average per account. +/- tolerance configurable. Spikes trigger review; flat-line suggests meter failure.

Account master reconciliation

Every bill matched to an account in your utility master. Orphans (no matching account) and phantoms (account with no recent bill) both surface.

Asynchronous batch

Month-end runs hundreds of bills in minutes. Processing is queued so normal UI stays responsive during peak cycles.

GL coding & ERP post

Utility charges coded to the correct expense accounts per property / cost center. Posted via the same Invoice connector as AP.

Flow

End-to-end, document to ERP.

1
Ingest
Email, SFTP, or provider portal. Classified by provider and bill type.
2
Extract
Account, period, meter reads, usage, amount, tiered charges.
3
Compare
Usage vs 12-month rolling average. Tolerance check.
4
Reconcile
Matched to utility account master. Orphans routed to ops.
5
Post
GL-coded and posted to ERP. Anomalies held for ops review.
Compliance

Built to the standard the data requires.

Every Clarity Utility deployment operates under the same controls shared across the Clarity platform — and the app-specific ones listed on the right. Field-level audit. Encryption everywhere. Principle of least privilege.

App-specific controls
  • Source retentionOriginal bill preserved with extraction
  • Audit logEvery read, every anomaly, every post
  • EncryptionAES-256 at rest, TLS 1.2+ in transit
  • Idempotent postingSafe to retry, no double entries
  • Per-provider isolationProvider-specific API credentials vaulted
Integrations

Plugs into what you already have.

Pre-built connectors for the common systems. File-based, SFTP, webhook, and portal-automation fallbacks for the rest.

Municipal utility billing systemsMeter data management (MDM) systemsCustomer information systems (CIS)MUNIS / Tyler FinancialsSAP ECC / S/4HANAOracle Cloud ERPWorkday FinancialsSFTP / email intake
Production proof

Already running in production.

97%
extraction accuracy

Measured against a hold-out set of real municipal utility bills across water, gas, electric, and sewer providers.

Minutes
per month-end batch

Hundreds of bills processed asynchronously. Normal UI stays responsive during peak processing cycles.

Zero
meter-failure misses

Flat-line detection catches stuck meters before the utility ever notices. Customer-reported meter issues dropped 80%.

Frequently asked

How does it handle tiered or demand-based pricing?
The extraction schema captures all billed line items including tier boundaries and demand charges. Validation rules check that the sum of tiered charges reconciles to the total, and that demand charges match the peak usage reading.
What if a provider changes their bill format?
The system notices format drift automatically — confidence scores drop when a schema no longer fits. You'll get a notification that the provider has changed formats and we adjust the schema (typically same-day during pilot, within SLA during production).
Can we process historical bills retroactively?
Yes. Batch ingest supports any backfill window. We often process the last 12 months on pilot start so the anomaly detector has a baseline from day one.

Ready to try Clarity Utility?

Launch the live demo with sample data, or talk to us about a proof-of-concept on your real documents.