Cambridge-based AI software company Undo has secured €31 million ($37 million) in fresh funding to accelerate product development and expand its global presence as demand grows for reliable AI-assisted software engineering tools.
The funding round was led by Elsewhere Partners and marks a significant milestone for Undo, a company focused on helping engineering teams identify and resolve complex software issues through AI-powered root cause analysis.
As AI-generated code becomes increasingly common across enterprise development environments, software teams face a growing challenge: understanding, debugging, and maintaining systems that are becoming more complex than ever before. Undo aims to solve this problem by providing the runtime visibility and context needed for both human engineers and AI agents to diagnose issues accurately.
Undo's Mission: Making AI-Generated Code Easier to Debug
Founded in 2012 by CEO Greg Law, Undo has spent years developing deterministic program recording technology that captures how software behaves during execution. This technology creates complete execution histories that allow developers and AI systems to analyze failures and identify root causes with significantly greater accuracy.
According to Law, the rise of AI-assisted development has made runtime visibility more important than ever.
"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 Greg Law, Founder and CEO of Undo.
The new investment will help the company integrate its technology deeper into AI-driven development workflows while expanding commercial operations across North America and Europe.
Why AI-Assisted Software Development Needs Better Visibility
The rapid adoption of AI coding assistants has increased developer productivity, but it has also introduced new challenges.
AI tools can generate large amounts of code quickly, yet engineers often struggle to fully understand, trust, or debug that code. As a result, software systems can become increasingly difficult to maintain, leading to reliability issues, security risks, and operational disruptions.
Elsewhere Partners Operating Partner Rod Favaron believes this challenge is becoming one of the biggest obstacles in modern software engineering.
"AI is making code unmanageable by introducing software that engineers cannot always understand, trust, or debug. Systems become filled with unknowns, increasing the risk of outages, security incidents, and customer escalations."
Undo addresses this challenge by providing the runtime context that AI models need to make informed decisions during debugging and root cause analysis.
How Undo Uses Runtime Context to Improve AI Performance
One of Undo's core beliefs is that AI performance depends heavily on context.
The company summarizes this with a simple formula:
AI = Model + Context
While large language models can analyze source code, they often lack visibility into what actually happened when a program was running. Runtime context bridges this gap by showing the real execution history rather than simply displaying static code.
Undo captures and stores complete execution recordings, enabling AI agents to investigate software behavior across development, testing, and production environments.
This additional context significantly improves the ability of AI systems to identify the underlying causes of software failures.
Benchmark Results Show Significant Improvements
According to company benchmarks involving complex software bugs, Undo's runtime context dramatically improves AI debugging performance.
Key findings include:
-
AI models identify the root cause of only 38% of complex issues without Undo's runtime data.
-
When provided with Undo's runtime context, root cause identification increases to 92%.
-
AI models require fewer tokens to solve problems when runtime context is available.
-
Customers report root cause analysis processes completing up to 100 times faster.
These results suggest that runtime intelligence could become a critical component of enterprise AI development workflows.
Enterprise Adoption and Customer Success
Large enterprises are already using Undo to improve software quality and reduce debugging time.
One notable customer is Palo Alto Networks, where engineering teams manage multi-million-line codebases that can be difficult to troubleshoot using traditional logging and monitoring tools.
Suresh Sangiah, Senior Vice President of Engineering at Palo Alto Networks, highlighted the importance of visibility into runtime behavior.
"The hardest and costliest bugs live in runtime state and are not captured by logs or other solutions. Undo provides the visibility needed to catch and correct errors before they become operational problems."
According to Sangiah, Undo can often identify root causes autonomously within minutes.
Growing Investment in AI Software Infrastructure
Undo's funding arrives amid strong investor interest in the infrastructure layer supporting AI-powered software development.
Throughout 2026, companies focused on software supply chain security, AI governance, cloud operations, developer tooling, and production reliability have continued to attract funding. Industry observers increasingly view operational visibility and reliability as essential components of the AI software stack.
Within the UK ecosystem, companies such as Cloudsmith and other AI infrastructure startups have also secured funding, reflecting growing demand for technologies that help organizations maintain control and reliability as AI adoption accelerates.
What's Next for Undo?
With €31 million in new capital, Undo plans to significantly expand its:
-
Product development teams
-
Customer success operations
-
Sales and go-to-market efforts
-
Presence across the United States and Europe
The company aims to position itself as a foundational layer for AI-assisted software engineering, ensuring that both human developers and AI agents have access to the runtime intelligence needed to understand, maintain, and improve increasingly complex software systems.
As enterprises continue integrating AI into software development workflows, tools that improve reliability, observability, and automated debugging are expected to become a critical part of the modern engineering stack. Undo's latest funding round suggests investors believe runtime context will play a central role in that future.
