Skip to content
Portrait of Peter Schrammel

The Luvo 100 / Global

98No. 98 of the
Luvo 100 Global

Peter Schrammel

Co-founder and Chief Technology Officer
Diffblue
United Kingdom

AI & Quality

Why Peter Schrammel is in the Luvo 100

Peter co-founded Diffblue and helped bring automated test generation into enterprise development workflows. The company’s tools support developers working across Java codebases, including older systems where regression tests help detect unintended changes. Research on software verification and evaluating AI-generated tests, alongside conference talks and writing, contributes to a more rigorous understanding of how automation can support software quality.

Peter Schrammel’s impact

Advanced AI testing, automated unit testing, program analysis through widely used ideas, tools, research, teaching or community leadership.

About

Peter Schrammel is Co-founder and Chief Technology Officer at Diffblue, a University of Oxford spin-out developing software for automated test generation. Founded with Daniel Kroening, the company connects research in analysing software with tools for development teams. Diffblue generates unit tests to help teams detect unwanted changes in behaviour, with a Testing Agent that coordinates test generation, build fixes and verification across codebases, including older Java projects.

Peter also conducts research in software verification, using mathematical methods to check whether programs behave as expected. Recent joint projects include Iekkë, a tool for checking C programs with multiple threads of execution, and RESTestBench, a benchmark for evaluating AI-generated tests of web interfaces against requirements written in everyday language. Conference speaking and writing extend these ideas into industry, including an invited talk on academic research and industrial impact alongside writing on engineering reliable AI agents.

Career highlights

  • Co-founded Diffblue with Daniel Kroening as a University of Oxford spin-out
  • Co-authored RESTestBench, accepted at EASE 2026, on evaluating AI-generated web-interface tests
  • Presented joint work on the Iekkë verification tool at ISSTA 2026
  • Delivered an invited BASiC talk on academic research and industrial impact in 2025
  • Served on the programme committee for ICSE 2027