The annual release of the Sunday Times Tech Track 100 is a welcome celebration of the UK technology scene at a time as the country looks to the sector as an increasingly important part of its future.
As nurturing UK tech companies moves higher up the agenda, perhaps the example of cyber security start-up Darktrace, an entry in the ‘One to Watch’ section, could provide lessons for start-ups, the security industry and government alike.
The company was founded in 2013 and now has 350 customers with 1500 deployments, with teams in 23 countries around the world. Its most recent funding round included investment from Samsung.
It was founded by three distinct groups, Dave Palmer, Director of Technology explains to CBR: mathematicians from Cambridge, software specialists and people who had worked with UK Government.
“There was no way of finding out what was going on in your environment if you didn’t know how to go and you were searching for,” explains Palmer.
“If you knew what you were looking for there were stacks of technical approaches you could use. But there was no way of being notified that something unusual was happening or something has changed.”
The approach used in the company’s product Antigena is based on advanced probability and machine learning.
Palmer explains that it combines a dozen machine learning approaches, each one with different strengths and weaknesses, that compete to best represent the user.
“We look for what is unusual and understanding the context of what it is rather than predefining what attacks look like and escalating that to a defender.”
A sales representative might commonly access certain folders on the network, but an expansion of their machine’s activity beyond this might suggest a cyber attack.
“There are 350 types of information and quite a lot of it,” says Palmer. “It’s not unusual to get a million pieces of information per second.
“We push that [information] into the dozen or so machine learning approaches.”
However, the approach is probabilistic rather than binary. Each machine learning approach produces a metric for how unusual they believe the activity is.
As the activity continues, the score increases as it becomes more and more certain that there’s a problem.
Organisations may assign different levels of risk to different departments. For example, activity involving a company’s intellectual property might be given high priority compared to other parts of the organisation.
“The probability theory is curating in line with what the business has said is relevant or not,” explains Palmer.
Of course, having an original or innovative technology proposition is not enough to turn a small team into a world-beating tech company. How has Darktrace secured its place on the Ones to Watch list and what do other UK tech companies need to do to get there?
“Most places where the emerging ideas are coming from have a big presence of American companies,” says Palmer.
“There are outstanding ideas and people who can do amazing things here in the UK, but we have so many attractive big companies here.”
For example, Microsoft, Google and IBM all have a presence near Darktrace’s Cambridge headquarters.
Emily Orton, Head of Marketing at Darktrace, says that the UK does have the needed skills, but that the aspiration is not there in terms of thinking big.
Part of this is in terms of the way UK tech companies project themselves.
“We tend to be not quite so good at marketing ourselves,” she says. “The British tech industry underplays itself.”
From a business standpoint, Palmer emphasises the importance of listening to customers.
“Don’t spend three years polishing your product before you begin to think about how to sell it and how customers will adopt it,” he says. “You will change your mind quite quickly between how you think customers want to work and what they actually want to do in the real world.”
For example, Darktrace believed that its first customers would be in banks, but soon found that there was considerable demand from power station Drax and transport operator Virgin Trains.
“That might not affect the underlying theory of what you want to do but may affect the mechanisms of how you build it.”
As for the ecosystem, Palmer says that the UK government has been highly supportive, with delegations to the US and India helping the company spread its message. However, he says that navigating the export market is difficult for UK tech start-ups.
Whether the UK can become as prolific and long-term a generator of tech start-ups as the US remains to be seen, but the future trajectories of companies such as Darktrace will provide some indication.