Recovery from records estates would otherwise abandon.

The problem

When a technology company fails, nobody thinks of its internal records as an asset. They're scattered across code repositories, ticketing systems, shared drives and chat tools, and no one left on the estate has the time or the background to go through them. Handling them carelessly can create privacy and liability problems of its own, so most estates just leave them alone.

What changed

That's starting to change. AI companies now license real engineering and operational records: the kind of material that shows how work actually got done inside a company. Records an estate would otherwise write off can be worth something to creditors instead.

What I do

• InventoryFigure out what actually exists: repositories, documentation, project history, internal communications.
• ScreenTrace where the material came from and who owns it, working with the estate's counsel. Flag anything held in licensed third-party systems, and identify personal information to remove or exclude under PIPEDA and provincial privacy law.
• Package & marketPut the records together and take them to buyers who are actually in the market for them.
• StructureWork with the estate's counsel on terms that keep the estate protected.

How I work

I act for the seller, not the buyer. I don't buy records.

A first conversation costs the estate nothing.

About

I'm Adam Crespi, an embedded software engineer based in Vancouver (UBC Engineering Physics). Going through a codebase and its history to work out what's actually there is engineering work, and it's the part most estates have no one to do.

Portfolio: adamcrespi.github.io

Contact