Confidential data licensing for operating businessesConfidential data licensing

Data licensing for operators

The records your business keeps are now an asset.

AI labs need real data on how companies sell, serve customers and get work done. You hold years of it in your CRM, your support desk, your finance system and your inbox. We value it, remove every name and identifier, and license it on terms you set.

Request a confidential valuation How the process works
License No. AF-2026-0417Class · Operational records

Certificate of Data License

Licensor
Fernhill & Co.COMPANY_014
Records
412,560 tickets · 38,902 deals · 96,118 invoices
Period
Jan 2019 to Jun 2026
Identifiers removed
2,961,447
Status
De-identified · Verified · Licensed
Licensee · Frontier AI lab
$25K–$300K+
indicative value of an operating history, plus a share of every license that follows.
Zero
raw records ever licensed. Only de-identified data leaves the pipeline.
One call
from your team. We handle export, de-identification, packaging and sale.

The asset

Every deal, ticket and invoice records a real decision.

A discount offered to save a renewal. A complaint escalated before it became a refund. A supplier pushed back on price. Models trained on the open web have never seen these decisions, because they live inside systems like yours.

Labs building agents for sales, support, finance and operations need exactly these records, and they pay to license them. Until now there was no one to value it, clean it and sell it on your behalf.

Schedule A · Eligible records

Records we value and license, by function

Sales and customers SC

  • SC-01CRM accounts, deals and pipeline history
  • SC-02Quotes, proposals and contracts
  • SC-03Pricing and discount decisions
  • SC-04Customer support tickets and chats
  • SC-05Email threads with customers

Finance FN

  • FN-01Invoices, payments and collections
  • FN-02Purchasing and vendor bills
  • FN-03Budgets, forecasts and plans
  • FN-04Expense and approval workflows
  • FN-05Ledger entries and month-end close

Operations OP

  • OP-01Orders, fulfilment and returns
  • OP-02Projects, tasks and tickets
  • OP-03Inventory, scheduling and capacity
  • OP-04Supplier and vendor threads
  • OP-05Internal documents, wikis and runbooks

Process

Four steps. Most of the work is ours.

Your team answers a few questions and grants access. We handle export, de-identification, packaging and sale.

  1. I

    Valuation

    Tell us which systems you run and how long you have run them. We return a written valuation.

    You · a short call

  2. II

    Offer and scope

    We present an offer: a payout plus a revenue share. You choose which systems and date ranges are included.

    You · review and sign

  3. III

    De-identification

    Records are exported into an isolated environment. Every person, company, address and account number is removed.

    Alderfly

  4. IV

    Payment

    You receive a payout for your historical records. New records follow the same path and continue to pay.

    Alderfly · recurring

De-identification

Names, addresses and account numbers never reach a licensee.

Business records are full of identifiers that ordinary tools miss: account and invoice numbers, customer names deep in an email thread, signatures on a scanned contract, bank details in a PDF invoice. Our pipeline is built for these documents.

What we remove

Every category below is removed or replaced with a consistent pseudonym, so relationships in the data survive while identities do not.

PeopleNames, emails, phones, signatures
CompaniesCustomers, suppliers, partners, vendors
PlacesStreet addresses, site and office IDs
Reference numbersAccount, invoice, order, ticket numbers
Financial detailsBank accounts, cards, tax IDs
CredentialsAPI keys, tokens, passwords

How the pipeline works

  1. A detection model trained on business documents flags identifiers in text, tables and attachments.
  2. Rule checks catch structured numbers by format, such as IBANs, card numbers and carrier tracking numbers.
  3. Identifiers are replaced with consistent pseudonyms, and dates are shifted uniformly to preserve timing.
  4. A reviewer samples every release before it is packaged for licensing.
From Daniel Price <daniel.price@calderrowe.com>PERSON_05 <EMAIL_05> To Maya Ortiz <maya@fernhillco.com>PERSON_02 <EMAIL_02> Subject Renewal and invoice INV-2024-0871

Hi MayaPERSON_02,

Our board approved the renewal this morning. We can sign the two-year term at 48,600 a year from 1 October, if the 7% discount still stands.

Please address the contract to Calder & Rowe LLP, 41 Bedford Row, LondonCOMPANY_031, ADDRESS_09. Payment will come from account GB29 NWBK 6016 1331 9268 19ACCOUNT_03, as before.

If legal needs more time, we can extend the current term by 30 days at the old rate.

Best,
Daniel PricePERSON_05, Operations Director
+44 20 7946 0321PHONE_05

Identifier removedSignal kept Sample record

Licensing policy

Who may license your data, and who may not.

Licensed to

AI labs, research teams and data partners, under a signed license.

  • Named licensees with audited use terms
  • De-identified records only
  • No raw records, ever
  • No re-identification attempts permitted
Never licensed to

Your competitors.

Who we work with

From first contact to final invoice.

If your business runs on software and has kept its records, there is something to license, whether you sell software, services or physical goods. The most valuable datasets follow a customer across several of these stages.

  1. Marketing

    Marketing and growth teams

    CampaignsInbound leadsWeb analytics

  2. Sales

    Sales and account teams

    PipelineProposalsContracts

  3. Delivery

    Operations, fulfilment and project teams

    OrdersProjectsSchedules

  4. Support

    Support and customer success teams

    TicketsChatsRenewals

  5. Finance

    Finance and procurement teams

    InvoicesPaymentsVendor bills

Contact us

This form is only for getting in touch with our team. To have your data valued, request a valuation.

Your details are used only to reply to you.