For PE sponsors, portfolio companies and owner-led businesses

Your portfolio is sitting on data AI labs will pay for.

AI labs pay for real operating records, the kind built up over years inside every portfolio company. We strip out anything that identifies a person and license the rest only with your approval.

  • You approve every buyer
  • Never sold to competitors
  • No capex, no disruption to management
See how it works

Free · Takes 2 minutes · Valuation in 24 hours

Fund IV · portfolio data inventorySample
HoldingHistory
Project Atlas
Home services
11 yrs
Project Meridian
Specialty healthcare
9 yrs
Project Forge
Industrial mfg.
14 yrs
Project Relay
Freight & logistics
8 yrs
Project Ledger
B2B software
7 yrs
Indicative value per company, by size and years of data$50K – $5M

What AI companies have paid for data

  • Reddit → Googleabout $60M a yearReuters
  • Reddit$203M in licensing contractsTechCrunch
  • News Corp → OpenAI$250M+ over five yearsPress Gazette
  • Tempus → AstraZeneca and Pathos$200M over three yearsTempus
  • Closed startups' work archives$10K to $100K eachForbes
  • AI training tasks$200 to $2,000 eachEpoch AI

AI models have read most of the public internet but almost none of the records businesses keep. Labs now pay for those records, and a typical PE portfolio holds years of them across every holding.

You decide what to share and who can buy it.

We do the legal and technical work and handle the sale. You approve the data before it's offered and each buyer before they sign. Run it for one company or across the whole fund.

  1. 01

    Start with a call

    A 30-minute conversation with the deal team or operating partner about which holdings, systems and years of history are in scope.

    We never need access to your systems

  2. 02

    We check you're allowed to sell

    We read customer contracts and privacy terms with management before any data moves. Anything you can’t license cleanly stays out.

    Mutual NDA from the first call

  3. 03

    We clean the data and you review it

    Our software replaces names, contact details and account numbers, deletes passwords, and rounds dates and amounts so no one stands out. You see exactly what a buyer would get.

    Nothing is offered until you sign off

  4. 04

    We find the buyer

    We show a cleaned sample to the AI labs and data suppliers you approve, and negotiate a license that limits how they can use it.

    Never sold to competitors

  5. 05

    You get paid

    The license pays for the company’s history. Proceeds flow to the portfolio company, with no capital required and no change to day-to-day operations.

    We’re paid only when you are

What the data is worth

$50K–$5M

Data can be valued from $50K to $5M per company, depending on company size and years of data. Across a portfolio, that adds up.

Company size
More customers, transactions and employees mean more records, and more variety in how work gets done.
Years of data
Long, consistent history is what labs prize most. A decade of records is worth far more than a year.

Names and contact details never reach a buyer.

Labs pay to learn how work happens: what was decided, in what order, and what came of it. They don't need to know who was involved.

Names
Replaced with the same code everywhere
Emails and phones
Removed
Account and project numbers
Replaced with codes
Amounts
Rounded
Dates
Cut to the month
Sites and addresses
Replaced with codes
Passwords and API keys
Deleted, and we tell you
Practice management · Claims · CLM-448120
Original

Dr. Maria Chen · Lakeside Orthopedics · Patient DOB 04/11/1968 · CPT 27447 · $38,214.60 · Jun 3, 2024 · Denied, appealed, paid

What a buyer sees

PROVIDER_0412 · SITE_018 · Patient age 55–60 · CPT 27447 · ~$38,000 · Jun 2024 · Denied, appealed, paid

Fields changed
5
Secrets left
0
Smallest group
k ≥ 5

Patient records are de-identified to HIPAA standards, either Safe Harbor removal of the 18 identifiers or Expert Determination, before anything is offered.

Labs pay most for work the internet doesn't show.

We focus on the sectors where private equity invests most, where buy-and-build platforms have accumulated years of detailed operating records.

  • Healthcare & healthcare services

    Physician practice groups, dental and vet, behavioral health, home health, RCM

    • De-identified claims and denials
    • Scheduling and utilization
    • Revenue cycle workflows
  • Business & financial services

    Insurance brokerage, specialty lending, accounting, staffing, legal services

    • Underwriting and claims files
    • Credit decisions and memos
    • Ledgers and close history
  • Industrials & manufacturing

    Specialty manufacturing, aerospace and defense suppliers, packaging, chemicals

    • Downtime and maintenance logs
    • Quality checks and CAPAs
    • Orders and scheduling
  • Software & technology

    Vertical SaaS, IT and managed services, tech-enabled services

    • Support tickets and resolutions
    • Product plans and decisions
    • Incident history
  • Consumer & home services

    HVAC, plumbing and electrical, pest control, landscaping, franchise systems

    • Job and dispatch histories
    • Quotes, pricing and upsells
    • Call recordings and transcripts
  • Infrastructure, energy & logistics

    Freight brokerage, 3PL, utilities services, environmental, field services

    • Load, lane and rate history
    • Asset inspection reports
    • Safety and compliance logs

If a portfolio company has kept detailed records for years, we can value them.

Built for sponsors

Works the way a fund works.

Whether you're a deal partner looking across the portfolio or a CEO at a single company, the process and the protections are the same.

One program, every holding
Operating partners can run a single review across the fund. We use the same contracts, the same anonymization and the same buyer rules for every company.
Value creation without capex
Licensing proceeds come from an asset already on the balance sheet. No new hires, no new systems, no change to the operating plan.
Light lift for management
A 30-minute call, a contract review with us, and a data export. We find buyers, negotiate and handle the paperwork.
Clean at exit
Every license is documented with usage limits, a no-resale clause and a buyer list you approved, ready for the next owner’s diligence.

Protections built in at every step.

  • Contracts checked first

    We read customer contracts and privacy terms before any data moves, and leave out anything you can’t license.

  • Personal details removed

    Names, emails, phone numbers, account numbers and addresses are replaced or deleted. Dates and amounts are rounded so no one stands out.

  • Passwords removed

    Passwords, API keys and other credentials are deleted, and the report lists them so you can change them.

  • Cleaned in your environment

    Records are scrubbed before they leave your network. A sample you share for review is cleaned in memory, and the original is never stored.

  • You approve every sale

    You see what a buyer would get and approve each buyer. You can say no at any point.

  • Never sold to competitors

    You list the companies that can never buy, including other sponsors’ portfolio companies. Every license bans reselling the data or trying to identify anyone in it.

Questions

Is this only for private equity owned companies?

No. Any executive with ownership power can reach out, whether the company is founder-owned, family-owned or backed by a sponsor, and potentially get massive value for the data.

Who buys the data?

AI labs that train models for business, science and industry, and the companies that supply them with data. You approve every buyer before a license is signed and can rule out any company.

Is this legal? Our customers trust us with their information.

It is if your contracts allow it, so we check them first. Data you hold for customers usually needs their written permission, and if we can’t get a clear answer we leave that data out.

What exactly leaves the company?

A cleaned dataset, after you review a sample of it. Names, contact details, account numbers and passwords are removed or replaced, and dates and amounts are rounded so no one can be picked out.

How much is the data worth?

Data can be valued from $50K to $5M per company, depending on company size and years of data. Share a few details and we send a valuation within 24 hours.

Can we run this across multiple portfolio companies?

Yes. Most sponsors start with one or two holdings, then extend the same contract review, anonymization and buyer rules across the fund.

How much work is it for management?

A 30-minute call, a contract review with us, and a data export. We find buyers, negotiate and handle the paperwork.

How do you get paid?

Only when a license pays. If nothing sells, you pay nothing.

Find out what your portfolio's data is worth.

Answer a few quick questions and we'll send a data valuation within 24 hours. We sign a mutual NDA before any data moves, and we never need access to your systems.

jsilver@dataassetgroup.com

See what your data is worth

Takes 2 minutes. We never need access to your systems.

1About you

Your role

2Systems you'd consider sharing

Systems

3Years of records

6 y

4How big is the company?

5Can you license this data?

Licensing rights

6Where should we send your valuation?

We won't share your details, and we sign a mutual NDA before any data moves.