Skip to content
Gainfluence

Data brokerage for AI training and evaluation

We license proprietary company data to AI labs.

We source data from operating and closed companies. We value it, package it and run the sale across many buyers. Sellers get paid for data they already have. Buyers get licensed, documented, model-ready datasets.

Sell your dataRequest catalog access

01What we broker

Every kind of proprietary data, through one broker.

Four main categories, and anything else that is hard to find elsewhere.

A

Workplace data

How real teams communicate, plan and get work done.

  • Email
  • Slack and Teams chat
  • Documents
  • Calendars
  • CRM (Salesforce, HubSpot)
  • Support tickets
  • Project trackers (Jira, Linear, ClickUp)

B

Source code

Full repositories with the context around them.

  • Complete repos with commit history
  • Linked issues and pull requests
  • Code review threads
  • Design files
  • Internal technical docs

C

Gameplay

What players did, and the games they did it in.

  • Gameplay footage
  • Player input logs
  • Telemetry and event data
  • Game builds
  • Engine projects (Unity, Unreal, others)

D

Industry data

Operating records from specialised sectors.

  • E-commerce
  • Real estate
  • Construction
  • Professional services
  • Healthcare operations
  • Media and film
  • Other sectors

E

Anything else that is hard to find

Scarcity drives value. If few others hold it, ask us.

02Process

Four steps from submission to payment.

We handle sourcing, valuation, packaging, the buyer process and delivery.

See the full process

  1. 01

    Submit

    Tell us what data you hold.

  2. 02

    Valuation

    Get a written valuation range.

  3. 03

    Buyer process

    We run a competitive sale.

  4. 04

    Paid

    Deliver the data and get paid.

03Why a broker

A competitive sale, not a single offer.

Selling direct means one buyer sets the price. A broker brings several buyers to the same dataset.

Selling direct compared with selling through Gainfluence
TopicSelling directThrough Gainfluence
BuyersOne buyer, on its own terms.A competitive process across AI labs and data companies.
Data typesOften a single category.Workplace, code, gameplay and industry data.
Company statusOften closed companies only.Operating, winding down or closed.
ValuationYou get the buyer’s offer.You get a written valuation before you commit.
PackagingYou export and clean the data.We export, deduplicate, redact and document it.
Other servicesSometimes tied to a shutdown package.None. We broker data and nothing else.

04FAQ

Common questions.

Read all questions

What drives the price of a dataset?

Five things matter most. Volume: how many messages, files, repos, tickets or hours of footage exist. History: how many years the records cover. Headcount: how many people produced the data. Usage: how heavily and how consistently the team used its tools. Scarcity: how hard the data is to find elsewhere.

Linked context adds value. Code with its issues, pull requests and the chat around them is worth more than code alone. Rare languages and specialised domains are worth more than common ones.

How long does a sale take?

It depends on the dataset and on how many buyers are in the process. The stages are: submission, valuation, buyer process, then delivery and payment. We give you an estimated timeline with the valuation.

TODO: Add typical durations for each stage once you are comfortable publishing them.

Can an operating company sell its data?

Yes. We work with operating companies, companies winding down, and companies that have already closed. Operating companies often license historical data or a defined slice of it, and keep running as before.

Is the license exclusive?

It can be either. A non-exclusive license lets the same dataset be licensed to several buyers. An exclusive license goes to one buyer and is priced higher. Scope and term are negotiated per deal. We will tell you which structure we expect to get the best result for your dataset.

What happens to employee and customer data?

Personal information is removed or redacted before delivery. That covers names, email addresses, phone numbers, account identifiers and other direct identifiers, using automated detection with human review. The exact treatment is agreed with the seller and documented for the buyer.

TODO: Confirm the redaction standard and any legal review step that applies.

For sellers

Find out what your data is worth.

Send us an inventory. We reply with a written valuation range. There is no obligation to sell.

Sell your data

For AI labs

See what is available.

Verified buyers get listing details, documentation and evaluation samples.

Request catalog access