The repo plus everything around it
Labs have plenty of public code. What they are buying now is the context: the issue that prompted a change, the review that pushed back, the support ticket that found the bug, the postmortem that explained it.
Why labs want it
Code sales from closing startups are now common enough that SimpleClosure told MIT Technology Review that 47 percent of its wind-down deals include a code or data sale. Operating companies can license the same thing without closing.
micro1 lists software and engineering as a data area, and Handshake AI lists software among its industries, naming issue histories and development workflows.
Where this is reported, as of October 2026. None of these companies is a SellMyData customer; they are the public record of what buyers are asking for.
- MIT Technology ReviewReports that code and workspace data are being bought from startups, and quotes SimpleClosure’s CEO saying 47 percent of its deals now include code or data sales. AI labs are buying data from dead startups to train their models.
- micro1Lists software and engineering as a data area: engineering workflows, bug tracking, product planning and technical knowledge bases. micro1 Company Data Partnerships.
- Handshake AILists software among its industries: technical documentation, issue histories and development workflows. Handshake AI data partnerships.
What sells in software and saas
The five data types →Messages
Engineering and product channels, incident threads, release coordination.
Slack · Discord · Gmail
Code
Repositories with commits, pull requests, reviews and the issues that prompted them. Linked tickets raise the value.
GitHub · GitLab · Bitbucket
Documents
Specs, RFCs, design docs, runbooks, postmortems and onboarding guides.
Notion · Confluence · Drive
Records
Pipelines, accounts and subscription history with customer identities removed. No payment card data.
HubSpot · Salesforce · Stripe
Tasks
Issue trackers and support queues from report to fix, with the whole thread intact.
Jira · Linear · Intercom
Tool names are the ones we see most in this industry. If yours isn’t listed, it almost certainly still counts: take inventory and we’ll tell you.
What stays out
- Secrets and credentialsAPI keys, tokens, certificates and environment files are stripped from repositories before anything is listed.
- Your customers’ dataProduction databases and anything your customers stored in your product are never in scope. The dataset is about how you build and run the product, not what users put in it.
- Third-party codeVendored dependencies and code under licenses that don’t allow it are removed during the rights check.
- HR and personal channelsDirect messages, HR channels, payroll files and anything about an individual employee stay out.
Full detail: what we remove and how rights are checked. We don’t buy consumer personal data, health records tied to people, or payment card data.
How it works
- STEP 01
Value it
Answer a few questions and get an estimated range. Nothing is shared yet.
- STEP 02
Share your exports
We walk you through exporting from each tool you choose, for the date range you choose. We never write to your systems.
- STEP 03
Approve the listing
We strip names, contacts and identifiers. You review the redacted preview and the price before any lab sees it.
- STEP 04
Get paid
When a lab buys a license, you get paid. We aim to pay within 7 days of the sale closing.
Today this is done by hand, with you. You export, we clean and prepare, you approve. How you stay in control.
Questions from software and saas
Does our code need to be good?
It needs to be real. Buyers want the history of how a team worked, including the messy parts. A repository with honest reviews and linked issues is worth more than a tidy one with no context.
We use open-source dependencies. Is that a problem?
No, as long as they are left out. We separate your code from vendored and licensed code during the rights check.
Can we keep our core IP out?
Yes. You choose which repositories, folders and date ranges are in scope. Many companies license internal tools, older products or support queues and keep the core product out.
Will a lab be able to identify our product?
Product names and domain names can be replaced with placeholders along with people’s names. Tell us what needs to be masked and we include it in the preview you approve.
Something else? Email micki@sellmydata.com.