Last year, Legal Week Intelligence, in association with Fulcrum GT, profiled 20 innovators driving change in the legal sector. We focused on what inspired these figures to shift their views and how this bred innovation. In 2017, we look at the innovations themselves.
The Top 20 Legal IT Innovations report aims to get to the heart of what innovation means by illustrating how new ways of doing things, large or small, local or global, have shaken up a sector often characterised as slow moving and resistant to change.
RAVN Systems delivers artificial intelligence (AI), search and knowledge management solutions across many sectors, including legal, real estate, finance and general corporates. But the biggest is legal: RAVN serves law firms such as Dentons and Linklaters, while other clients range from British Telecom to the UK’s Serious Fraud Office.
The RAVN website describes “a tectonic shift in perceptions concerning applied AI systems.” Since the company was founded in 2010 by Peter Wallqvist, Van Hoecke, Sjoerd Smeets and Simon Pecovnik, perceptions about AI have certainly shifted. The quartet, none of whom are lawyers, met while working for Autonomy.
“There are two stories,” says Wallqvist (pictured), RAVN’s CEO. “One is technical – why these things are now possible – and the other is the credence that people put in the systems’ output.” Technically, he concedes that most core AI algorithms are not new – the theory behind them has been around for decades – but it is the real world application that makes the difference.
“The reason it is happening now,” he suggests “and probably even more so in the next few years, is that the sheer computing power to do some of these things wasn’t as widespread as it is now.” Wallqvist also points to “the sheer interconnectedness and digitisation of data – especially in the area where we operate with large document volumes and unstructured data.”
The main tectonic shift, he explains, is that the deviation between what a machine can do in the areas that RAVN operates in, and the correlation between the accuracy of the machine and the person, is essentially the same as between two people. When that can be statistically proven, perceptions shift.
Law firms are finding ways to exploit that efficiency gain, rather than thinking: ‘there goes my hourly billing targets’
Nevertheless, RAVN has a marketing issue. Wallqvist explains: “When people ask me to be concise, it’s a genuine problem. We have a core AI platform that is extremely capable of interrogating any sort of data source: document systems inside law firms, or CRM systems in professional services firms, for instance. So we have had to also develop end-to-end applications that exploit the capability of that platform for different, specific business reasons.
“Most of our clients are still law firms – if they have made the choice to become more efficient. It’s this inefficiency paradox that we have to overcome. The UK may be the most modern market in the world for legal services: we have the Law Society and the SRA to thank for that: they have forced law firms to think about being more efficient.”
He says that when RAVN tells some law firms: “This will take about a quarter of the time it took you before, it’s going to be more accurate and a lot quicker”, they just want to show you the door because they think that you’re making them poorer. They’re so wedded to the billable hour. Law firms adopting RAVN are finding ways to exploit that efficiency gain, rather than thinking: “there goes my hourly billing targets.’”
Market forces, he believes, will change minds.
RAVN has just passed the 50 staff headcount. In addition to London, there is a small Amsterdam office. “We’re growing quite a lot,” says Wallqvist. “A constant effort is to make our technology more consumable. The platform is now much easier for people that have had a little bit of training to use via an interface.
“When I speak to my counterparts in other tech companies in East London, they focus more on flashy web programming and apps, whereas we are engineers here. Maybe we wish we could be cool web people, but we’re grounded in a more mathematical algorithmic approach.”