Skip to content

Early access · Electricity bill reconciliation

Know what every electricity bill should have cost.

Wattwise AI reads your utility bills and meter data, checks every charge against your tariff, and flags overcharges, duplicate invoices and unusual usage — each with the numbers that prove it.

Built for hotels, retail chains and commercial property portfolios that pay many utility bills every month.

Harbor View Hotel · INV-HV-260831

August 2026 statement

Sample data

Billed consumption

48,920 kWh

Meter record

45,310 kWh

Energy
$6,946.64 $6,434.02
Demand
$2,275.00
Fixed charge
$45.00
Tax
$556.00 $525.24
Expected total$9,279.26

Billed $9,822.64 — $543.38 above expected

Billed kWh exceeds the meter record. Rates, demand and fixed charges match tariff GS-2.

Bill-to-meter reconciliation

Billed kWh is compared with your own meter records for the same period, and the cost of any gap is calculated.

Tariff-checked charges

Energy, demand, fixed and tax lines are recomputed from your configured tariff and compared line by line.

Duplicate invoice detection

Re-sent and re-uploaded bills are matched on invoice number, account, period and amount before they are paid twice.

Explained exceptions

Every flag comes with a plain-language explanation written from the calculated figures, not estimated by a model.

The problem

Utility bills are paid on trust because checking them does not scale.

Multi-site operators pay electricity bills every month across many accounts, tariffs and meters. Verifying each one by hand costs more time than most teams have, so the bills mostly get paid as issued.

Bills arrive in every shape

A hotel group with dozens of accounts receives PDFs from several utilities, each laid out differently, plus meter exports from building systems that never meet them.

Checks happen after payment

Accounts payable pays on the due date. Spot checks happen, if at all, when someone has time before month-end close, which rarely covers every site.

Errors are small and recurring

A wrong rate after a tariff change, a re-sent invoice paid twice, or kWh billed above what the meter recorded can repeat every month before anyone notices.

Manual review

  • Open each PDF and key figures into a spreadsheet
  • Compare to meter data only when something looks wrong
  • Find duplicates by memory or by search
  • Dispute with screenshots and email threads

With Wattwise AI

  • Fields are extracted from each bill with confidence shown
  • Every bill is compared with its meter record automatically
  • Duplicates are caught on upload, before payment
  • Disputes start from a calculated variance and the source files

The principle: code does the arithmetic, AI explains it. Every variance is calculated deterministically from the bill, the meter record and your tariff. Language models are used to describe those results, never to produce them.

Features

Everything a bill review needs, in one workflow.

Each capability feeds the next: extracted fields are checked against meter data and tariffs, exceptions are explained, and reviewers close them out with a record.

Bill ingestion & field extraction

Upload PDF bills as they arrive. Billing period, consumption, rates, demand and totals are extracted with a confidence score, and every figure stays linked to the source file.

No re-keying, and every number can be traced to its page.

CSV meter-reading import

Import register or interval reads from meters and building systems, mapped to the right site and account.

Your own data becomes the reference for each bill.

Tariff rule checks

Energy, demand, fixed and tax lines are recomputed from your configured rate schedule and compared with what was billed.

Wrong rates surface the month they appear.

Duplicate invoice detection

Invoice number, account, period and amount are matched against everything already recorded.

Re-sent bills are held before payment.

Unusual consumption flags

Usage is compared with each meter’s trailing 12-month average against a threshold you set.

Faults and leaks show up as bill lines, not surprises.

Bill-to-meter reconciliation

Billed kWh is set against metered kWh for the same period across all sites, and the cost of each gap is calculated.

Prioritise the sites where the money is.

Review notes & exception reports

Assign findings, record what you learned, and export an exception report for month-end close or a utility dispute.

Finance, facilities and property ops work from one record.

How it works

From a PDF in the inbox to a documented decision.

Five steps take a bill from arrival to a resolved finding. Select a step to see what happens.

Step 1 of 5

Submit records

Drop in PDF bills as they arrive and import CSV meter readings for the same sites. Originals are kept so every figure can be traced back.

  • harborview_aug2026.pdfReceived
  • hv_meter_aug2026.csvReceived
  • northgate14_aug2026.pdfReceived

Product

Try the workspace.

A working preview of the reconciliation workspace. Search and filter bills, open a finding to see the bill beside its meter record, change its status, add notes and export a report. All records are fictional sample data and changes stay in your browser.

Wattwise AI · Energy bills · Sep–Oct 2026

Sample data

Bills processed

6

3 sites · Jul–Aug

Flagged

4

Needs a decision

Unresolved

2

Open or in review

Amount in question

$1,581.34

Unresolved overcharges

Billed vs metered consumption (kWh)

Click a bar to open
0k40k80kHarborAugHarborJulNorthgateAugNorthgateAug (dup)ElmAugElmJul
MeteredBilledBilled above meter

Hover or focus a bar for detail

Activity

Largest unresolved finding

Elm Street Commons · Aug 1 – Aug 31, 2026 · +$1,037.96

Consumption matches the meter (61,200 kWh), but energy was billed at $0.1580/kWh instead of the configured GS-2 rate of $0.1420/kWh — $0.0160 higher. Applied to the full consumption and tax, the difference is $1,037.96. Check whether a tariff change notice was issued for this account.

Who it’s for

Built for teams that pay many electricity bills.

Hotel & hospitality groups

The problem
Dozens of properties, each with its own utility account, tariff and seasonal swings in usage.
How they use it
Reconcile each property’s monthly bill with its meter export and send the finance team only the exceptions.
The benefit
Overcharges are disputed with evidence while the billing period is still recent.

Multi-site retail & grocery

The problem
Many small accounts where individually minor errors, repeated across stores, add up.
How they use it
Check every store’s rate against the tariff and catch re-sent invoices before accounts payable runs.
The benefit
Duplicate payments and rate errors stop being found months later.

Commercial real estate owners

The problem
Landlord and tenant meters, shared services and pass-through charges that must be verified before recharging.
How they use it
Compare billed consumption with sub-meter records and flag usage spikes against each meter’s history.
The benefit
Property teams see equipment faults as unusual consumption, and recharges rest on checked figures.

Finance & shared-service teams

The problem
Month-end close leaves little time to question utility invoices across entities.
How they use it
Work a single review queue with owners, notes and an exportable exception report.
The benefit
Close packages show what was checked, what was disputed, and why.

Technology

How We Use Amazon Web Services

Wattwise AI turns a PDF bill into a checked, explained finding in five stages, and each stage maps to an AWS service. Files land in Amazon S3, Textract reads the fields, Lambda does the tariff math, Bedrock writes the explanation, and Cognito and API Gateway put the result in front of reviewers. This is the architecture the product is built around; the marketing site itself is static.

  1. 01

    Data collection

    Bills and meter files enter through authenticated upload endpoints.

    • Amazon API GatewayUpload and application endpoints
    • Amazon S3Receives original files
  2. 02

    Secure storage

    Source files and structured records are stored separately.

    • Amazon S3Original bills and meter exports
    • Amazon DynamoDBSites, tariffs, bills, review history
  3. 03

    Processing

    Fields are extracted, then reconciled in code.

    • Amazon TextractExtracts fields from bill PDFs
    • AWS LambdaTariff math and reconciliation
  4. 04

    AI intelligence

    A model explains results the code has already calculated.

    • Amazon BedrockPlain-language explanations
  5. 05

    Application

    Reviewers sign in, work the queue and receive alerts.

    • Amazon CognitoSign-in and user access
    • Amazon SESReview notifications
    • Amazon CloudWatchLogs, metrics, alarms

Services and what each does

  • Amazon S3: Stores original PDF bills and CSV meter files so every figure stays traceable to its source.
  • Amazon Textract: Extracts billing period, consumption, rates, demand and totals from bill PDFs, with confidence scores.
  • AWS Lambda: Runs the deterministic tariff calculations, duplicate matching and bill-to-meter reconciliation.
  • Amazon DynamoDB: Holds sites, tariff configurations, extracted bills, findings and review history.
  • Amazon Bedrock: Writes plain-language explanations from calculated results; it does not perform the arithmetic.
  • Amazon API Gateway: Fronts the upload, query and review endpoints used by the application and the website forms.
  • Amazon Cognito: Handles authentication and role-based access for finance, facilities and property users.
  • Amazon SES: Sends review assignments and delivers early-access and contact submissions to the team inbox.
  • Amazon CloudWatch: Collects logs, metrics and alarms for extraction, reconciliation and API health.

Business progress & early access

Help shape the first release.

Wattwise AI is an early-stage company building bill reconciliation for multi-site operators. We are working with teams who pay utility bills across many sites to refine extraction, tariff rules and the review workflow.

  1. 2026

    Wattwise AI founded

  2. Now

    Interactive product preview published, built on sample data

  3. Now

    Early-access requests open to multi-site operators

What to expect

  • We read every request and reply by email.
  • A walkthrough of the workspace using your bill formats and tariff types.
  • No payment details are requested at this stage.

Request early access

Fields marked * are required.

Optional. Tell us about your sites, utilities, or bill formats.

Pricing

Priced to the size of your portfolio.

Pricing depends on the number of sites and the volume of bills. We share it directly during early-access conversations.

Starter

A small portfolio proving reconciliation on its own bills.

  • PDF bill and CSV meter uploads
  • Field extraction with confidence scores
  • Bill-to-meter reconciliation
  • Duplicate and rate checks
  • Exception reports

Growth

Teams running month-end energy review across many sites.

  • Everything in Starter
  • Multiple tariffs and utilities
  • Review assignments and notes
  • Consumption trend flags
  • Help onboarding new bill formats

Enterprise

Multi-brand groups and large property owners with custom needs.

  • Everything in Growth
  • Single sign-on
  • Custom tariff configuration support
  • Security review and procurement support
  • Dedicated onboarding

Team

The Team Behind Wattwise AI

Wattwise AI is a small, founder-led company working on one problem: making electricity bills verifiable at portfolio scale.

A founding team that works directly with customers

During early access, the people building Wattwise AI are the people you talk to. They set up your first reconciliation, learn your bill formats and tariffs, and decide what is built next based on what your finance and facilities teams need.

Contact

Talk to the team.

Questions about the product, partnerships, or how we handle data? Send a message or write to us directly.