Undo raises $37M Series C led by Elsewhere Partners
What's the deal? Undo, a UK-based software debugging startup founded in 2005, has raised $37M in a Series C round led by Elsewhere PartnersDealroom has a profile for this one. Try Dealroom →. The company records the full execution history of a running program and saves it to a single file — capturing not what code looks like, but what it actually did when it ran.
The funding will go toward hiring across product, support, and sales teams in the US and Europe.
Undo sells its technology to engineering teams at networking, database, semiconductor, and financial services firms. Palo Alto Networks is among its customers.
Why now? AI coding assistants have changed the equation. Teams ship far more code than they can fully understand, and Undo argues that AI agents cannot reliably diagnose failures without seeing how software behaved at runtime.
The company's own benchmarks illustrate the gap. On a set of complex bugs, leading AI models identified the root cause 38% of the time on their own — a figure that rose to 92% when fed Undo's runtime recordings. Customers report completing root-cause analysis up to 100 times faster.
"Undo has spent years building deterministic program recording technology for code failure runtime visibility, which has become absolutely essential with the rise of AI," said founder and chief executive Greg Law.
What could go wrong? Undo's benchmarks are self-reported, and the company operates in a crowded observability market where larger players could build or acquire similar capabilities. Recording full execution histories also raises questions about performance overhead and storage at scale — challenges that grow as AI-generated codebases expand.
The signal: Elsewhere Partners' involvement is notable — the Austin-based fund focuses specifically on growth-stage software companies, having backed 18 firms across North America, Europe, and Israel since 2016. For a 21-year-old company like Undo, the round suggests patient capital betting that a long-standing technical moat is finally meeting its market moment, as AI-generated code creates urgent new demand for runtime debugging infrastructure.
Read more: SiliconAngle
