Let's talk
Finance and accounting

Fairbooks

The friend that helps you close. It writes the workpaper. It does not post, pay, or approve go-live.

A ledger page with checked rows, a bar chart and a stack of coins

Engagements

$4,500 · One campaign

Access review

Reviewers, due dates, segregation of duties, and the evidence pack.

$4,500 · One policy

Dispute packs

Reason codes, evidence, and the queue. A refund stays a person's click.

$7,500 · One migration

Migration readiness

Trial-load tie-out, duplicate masters, and the fix list for the next load.

$12,000 · One close

Month-end

Bank reconciliation, three-way match, expenses, and 1099 readiness.

Setup is in English. The checks read files from the United States, Germany, France, Spain, Brazil, Italy, Japan, China, and Korea, including the encodings those systems write. The finding sentences stay in English. This is not a statement about any country's law.

Thirty checks.

Close
Close

Bank reconciliation

Matches the statement to the ledger, including CAMT, BAI2, OFX, and MT940.

Close

Three-way match

Matches orders, receipts, and invoices, including partials.

Close

Expense audit

Reads receipt text and ranks what a person should review.

Close

Vendor readiness

Masked tax ids and who looks ready for a 1099.

Close

Dispute packs

Classifies a card dispute and lists the evidence. It does not refund.

Records
Records

Migration check

Ties a trial load by company and currency before go-live.

Records

Master records

Finds likely duplicate vendors and items for a person to merge.

Records

Inventory counts

Plans the count and shows the variance.

Records

Grant pack

Compares budget with actual and packs the evidence.

Records

Board figures

Checks a percentage against the locked number.

Access
Access

Access review

Reviewers, due dates, and segregation of duties.

Access

Tool review

Reads tool descriptions and lockfiles. It does not deploy.

Access

Certificates

Tracks expiry from the inventory you export.

Access

Control evidence

Maps the files you have to your own control list.

Access

Drift

Compares exported state with the approved baseline.

Papers
Papers

Clause review

Reads a contract against your playbook. It does not sign.

Papers

Erasure list

Applies your retention rules. It does not delete.

Papers

Consent scan

Compares a site capture with your vendor list.

Papers

License review

Reads lockfiles and flags a weak license for a person.

People
People

Onboarding list

The role checklist and the offboarding mirror.

People

RFI log

Ages the ball-in-court list from the log you export.

People

Seat review

Unused seats against the next renewal you can still cancel.

People

Script inventory

Estimates a conversion from the scripts you point at.

People

Doc drift

Checks a changelog against the version you name.

Quality
Quality

Eval review

Grades recorded judgments against your rubric.

Quality

Flow review

Reads an exported automation for secrets. It does not run it.

Quality

Send schedule

Holds a draft that misses a window. It does not send.

Quality

List hygiene

Bad addresses, role mailboxes, and withdrawn consent.

Quality

Mail auth

Reads DMARC and a saved header.

Quality

Page check

Contrast, focus, and labels from the static page.

Each one does one job, on your computer, from a file you export. There is no login to your bank, payer, carrier, or property system.

Measured in the final pass, 19 Nov 2026: 7 of the 30 checks

Measured, not promised.

44 of 44planted problems found
0false alarms on 82 clean rows
18 of 18one-change tests caught
4public data sources run

On fictional test data, 19 Nov 2026, for the 7 agents behind the four engagements (bank reconciliation, three-way match, expenses, access review, refunds and chargebacks, migration tie-out, duplicate vendors) of 30. The clean sets were written by the same author as the rules. The 1099 readiness agent is not measured: a clean vendor needs a well-formed taxpayer id, and the test files use invalid placeholders on purpose. The other 23 agents are checked by their eval cases only (596 expected markers, all found). This pass also fixed two bank-file reading bugs found on public files, two file-reading bugs found by the eight-language check, and one place where full tax ids reached the AI model. The test files and the checks were written by the same team, so these numbers show the rules do what they say. How often they are right on your own files is measured at your setup.

Run it yourself

git clone https://github.com/fpavia111-creator/fairbooks-agents
cd fairbooks-agents
npm install
node packages/fairbooks-bank-reconciliation/bin/fairbooks-bank-reconciliation.js demo

Open source, MIT licence, on GitHub: fpavia111-creator/fairbooks-agents. Needs Node.js 20 or later. The demo runs on made-up files and needs no network and no setup.

Common problems it finds

  • Bank fees and interest not booked, and checks and deposits not yet cleared, listed as reconciling items with a draft adjusting entry.
  • Invoices billed above the receipt or the PO price, duplicate invoices, invoices with no PO or no receipt.
  • Expense lines over the policy limit, with no receipt, sent late, without pre-approval, or claimed twice.
  • Vendors paid over the 1099 threshold with no W-9, an unsigned or old W-9, or a malformed taxpayer id.
  • People who left the company but still have access, admin accounts without MFA, shared accounts and service accounts with no owner.
  • Refund requests outside the return window, over the refundable amount, duplicated, or for an order that does not exist.
  • Migration loads where records are missing, amounts changed or totals do not tie with the old system.
  • Duplicate vendor records that share a tax id, phone or address.

Run on public data

SourceWhat ran and what it foundLicence
PostFinance sample files for business customers (camt.053)
2026-11-19
fairbooks-bank-reconciliation read all 4 statements; lines, totals and balances equal a separate re-count. In 2 of them the booked closing balance leaves out a 1,537.00 credit value-dated the next day; the agent now says so instead of only "does not tie".Published by PostFinance as sample files for software tests (no licence stated)
ofxparse test files (OFX bank and card statements)
2026-11-19
All 8 statements read; line counts, totals and ledger balances equal a separate re-count. No wrong results.MIT
bai2 test files (Citi, NatWest and SVB layouts)
2026-11-19
All 4 files read and equal a re-count. The SVB file holds two days for one account and did not tie because the reader took the second day's opening balance; fixed, it ties. The Citi sample's own balances do not add up, which the agent reports.MIT
SEC EDGAR XBRL company concept data (Apple, Microsoft, Walmart, Alphabet)
2026-11-19
fairbooks-erp-migration compared each period's first-filed value with its last-filed value: 448 periods, 26 reported differently later (for example Apple's fiscal 2008 net income, 4,834 then 6,119 million). The agent and a re-count list the same periods in all five series.U.S. government data, public domain

Public, lawful files only; no private data was used. The files are not copied into the repository: a script there downloads them again.

What a person still decides

PackageWhat a person reviews and decides
Access reviewEach system owner decides keep or revoke for every flagged account; revoking access is a one-click request a named person approves, and the agents never change access themselves.
Dispute packsA person reviews the reason code, the evidence list and the response draft before anything is submitted; a refund is always a person's click.
Migration readinessThe controller and the implementation lead decide whether untied totals, missing records and duplicate masters block go-live; go-live approval is never given by an agent.
Month-endThe accountant reviews matches, reconciling items, match exceptions, expense recommendations and the 1099 list, and posts any adjusting entry in the books; nothing is posted, paid or filed by the agents.
Export formats (44): 4 tested on a real or public file, 40 not yet tested on a real export.
SystemReport or exportStatus
OktaUsers API JSON (GET /api/v1/users)Not yet tested on a real export
Microsoft Entra IDUsers > Download users (CSV)Not yet tested on a real export
Microsoft 365 admin centerActive users > Export users (CSV)Not yet tested on a real export
Microsoft 365 usage reportsActive users report detail (CSV)Not yet tested on a real export
Google Workspace AdminUsers > Download users (CSV)Not yet tested on a real export
SlackManage members > Export member list (CSV)Not yet tested on a real export
GitHubOrganization People > Export (CSV or JSON)Not yet tested on a real export
GitHub CopilotSeat assignments API JSON (GET /orgs/{org}/copilot/billing/seats)Not yet tested on a real export
AWS IAMCredential report (CSV)Not yet tested on a real export
Atlassian AdministrationDirectory > Users > Export users (CSV)Not yet tested on a real export
BambooHREmployee report (CSV or Excel)Not yet tested on a real export
NetSuiteVendor saved search (default column labels)Not yet tested on a real export
NetSuiteTransaction saved search (default column labels)Not yet tested on a real export
NetSuiteItem saved search with location quantitiesNot yet tested on a real export
QuickBooks OnlineVendor Contact List report (CSV or Excel)Not yet tested on a real export
QuickBooks OnlineTransaction List by Vendor report (CSV or Excel)Not yet tested on a real export
QuickBooks OnlineTransaction Detail by Account / General Ledger reportNot yet tested on a real export
XeroContacts > Export (CSV)Not yet tested on a real export
XeroBills > Export (CSV)Not yet tested on a real export
XeroAccount Transactions report (Excel or CSV)Not yet tested on a real export
XeroBank statement CSV (Xero import layout)Not yet tested on a real export
StripePayments > Export (CSV)Not yet tested on a real export
ShopifyOrders > Export (CSV)Not yet tested on a real export
ShopifyProducts > Inventory > Export (CSV)Not yet tested on a real export
HubSpotContacts > Export (CSV or Excel)Not yet tested on a real export
SalesforceContacts or Leads report exportNot yet tested on a real export
SalesforceData Loader / Data Export of Contact (API field names)Not yet tested on a real export
MailchimpAudience > Export audience (CSV)Not yet tested on a real export
ProcoreRFIs > Export > CSVNot yet tested on a real export
RampTransactions > Export (CSV)Not yet tested on a real export
ExpensifyDefault CSV export of expensesNot yet tested on a real export
ChaseChecking or savings account activity (CSV)Not yet tested on a real export
ChaseCredit card activity (CSV)Not yet tested on a real export
Bank of AmericaChecking account activity (CSV, with its summary block)Not yet tested on a real export
Wells FargoAccount activity download (CSV with no header row)Not yet tested on a real export
Capital One360 account transactions (CSV)Not yet tested on a real export
MercuryTransactions export (CSV)Not yet tested on a real export
AWS Certificate Manageraws acm list-certificates JSONNot yet tested on a real export
crt.shCertificate search JSON (output=json)Not yet tested on a real export
ISO 20022 bank statementcamt.053 (PostFinance published samples, 2009, 2013 and 2019 versions)Tested on a public sample
OFX bank and card statementsOFX 1.x and 2.x statement downloadTested on a public sample
BAI2 cash management fileBAI2 prior-day file (Citi, NatWest and SVB layouts)Tested on a public sample
SWIFT MT940 statementMT940 customer statementNot yet tested on a real export
SEC EDGAR XBRLCompany concept data, converted to the migration compare columns by scripts/realdata.mjsTested on a public sample

No layout has yet been tested on a customer's own export. Every setup starts by confirming the layout on your files, and any column can be mapped by hand.

Security

What reaches the AI model

When an assistant calls a tool, its model sees the result: amounts, dates, ids, finding codes and text from your files, including people's names and e-mail addresses (Fairbooks has no general masking, unlike the newer Fair families). Masked: taxpayer ids (last four digits), secrets found in configuration and automation exports, card numbers in receipt text, and the person in an erasure request. The report and draft files written on your computer hold the full values. Results go to whatever model your assistant uses; for staff, customer and vendor data, use a model covered by your own agreements or a local model.

Support

Who keeps the rules current

Rules that change are settings or dated files, and a result names what it used: the 1099 threshold by tax year (600 before 2026, 2,000 in 2026, later years marked for confirmation), taxpayer id patterns, the cookie and tracker vendor list (version 2026.11.0), the MCP package advisory snapshot (2026-11-15), license text fingerprints and certificate strength rules. Expense, refund, tolerance, retention and access policies are your own JSON files, checked against a schema. You edit those files yourself, or support updates the bundled lists when a rule or a vendor changes.

How a setup runs.

1

You export the files

We send the list of reports by name. You export them from the systems you already use; nothing is connected.

2

Setup on your computer

Over a shared screen: the checks are installed, your columns are matched, and your own rules and dates are set.

3

The first report

The checks run on your files. We go through the worklist with you, and every flagged row is a person's call.

4

After that

You run them whenever you like. Support keeps the dated rules and the export layouts current.

A fixed scope and a fixed price, agreed before any work starts. Your files stay on your computer the whole time.

A sample report.

Written by the check itself on made-up files. Every person, business and number is invented.

Access review evidence All samples

The first page of the Access review evidence sample report