Public Funding for Tessian Limited

The world’s technological capacity to store information has exploded over the past 30 years to the equivalent of 4,500 stacks of printed books reaching from the earth to the sun. With the advent of internet related communications such as email, messaging, and online file storage, information security presents a completely new set of challenges - 76% of IT practitioners say their organisation experienced the loss or theft of company data over the past two years. Data Loss Prevention (DLP) tools are the answer. They aim to prevent the leakage of sensitive information via either accidental human error or malicious employee behaviour. Our company have developed CheckRecipient, the world’s first software platform to detect when emails are being sent to the wrong people, which is now in use with world-leading companies in the legal and financial industry. Our current software uses machine learning and artificial intelligence to analyse email data to classify the sensitivity of the data and the kind of recipients associated with this information. This is used to warn users about potential email related data loss before it happens. This project builds on our early technology and expertise to develop the world’s first fully-automated DLP platform to classify and protect all of an enterprises information from all potential forms of data leakage.

For most organisations email is the main artery of communication and a channel across which highly sensitive information is communicated and shared. Email is a highly vulnerable facet of an enterprise’s overall information infrastructure given the frequency and ease by which workers send emails every day and entire cyber security initiatives can be rendered redundant by a simple misaddressing error. Founded by a team of Imperial College trained engineers, ex-Investment Bankers and current finalists of the 3D Fintech Challenge 2015, CheckRecipient is an email security platform used by world-leading organisations to prevent confidential information being sent to the wrong person. Currently, there are two modules, CheckRecipient RuleBuilder, which allows organisations to design and implement customised, rule-based email communication policies and CheckRecipient AI, which performs a historical analysis of a sender’s email account to learn sending patterns and predict when an incorrect recipient may have been copied in on an email by mistake. Having developed and sold both of these products to a set of early customers, we are now looking to substantially transform CheckRecipient AI by incorporating natural language processing and machine learning technologies applied to the textual content present in email data to improve both the accuracy of predicting misaddressing errors and also develop the functionality of the product to protect against other email security risks. There is a strong market demand for an email security platform that is both able to work autonomously and with minimal disruption to the end user. Currently there exist a number of platforms on the market that look for predefined, specific text patterns (such as presence of social security numbers), but there are no products currently available that are analysing complex and unstructured textual content in emails to infer meaning and determine sensitivity and appropriateness for a set of recipients.