Posted: 15th December 2017

Project: Semmle
Official Title: Semmle: a powerful query language for analysing large data sources

Semmle is a successful spin-out company set up by Oege de Moor and employing a number of ex-members of the Department of Computer Science, based on their research on program analysis. Semmle markets an industrial-strength product allowing organisations with large software systems to understand and manage their code bases. This business intelligence platform started to be sold to prominent customers in 2008, including a multinational financial services holding company and NASA. NASA used it to help ensure the safe landing of the Curiosity Mars Rover in 2012.

Semmle was founded in December 2006 to create from scratch the novel technology that realises the potential of the research at Oxford in the context of Datalog, widening the scope of application to business intelligence rather than just program analysis. Further research was done in Semmle and six patents were filed by Semmle to protect these further advances after creation of the company.

Large software engineering organisations struggle with a lack of visibility of how outsourced development teams are performing. To create such visibility it is necessary to analyse many different data sources, including the code itself, version history, the bug database and test results. Much of this data are graphs and hierarchies, and thus writing the appropriate analysis to get visibility is hard. The research described here has (through further innovations at Semmle) led to a query language that, for the first time, opens up all the relevant data. Semmle's technology is routinely used in software production and management at numerous customers around the globe and has had significant benefits for these clients which include NASA, Murex, , Certipost, EMC, and a multinational financial services holding company.