iKVA has embarked on a £2.5m pre-Series A funding round to accelerate the development of its AI-powered knowledge management software solutions.
The “genesis” of iKVA in 2018 involved three academics from the university’s Department of Computing and Technology and the Alan Turing Institute – Dr Liang Wang, Professor Richard Mortier and Professor Jon Crowcroft – plus Jon , whose history includes stints at Cambridge Assessment and Local Heroes, a digital innovation from British Gas.
The latest funding round – “a pre-Series A round is also known as seed-plus” – follows a £1.5m seed investment raised exactly a year ago . iKVA has now raised a total of £4.4m in funding in four rounds from investors including Cambridge Enterprise and BTF.
iKVA’s cloud-based technology works by bringing together unstructured data from multiple internal and external sources – including rich documents, images, videos and chats – to deliver accurate information to the user in time. real as part of its regular workflow.
The startup’s platform uses vector analytics, which CEO Jon Horden describes as being context-driven rather than keyword-driven.
“We use vector mapping technology,” he explains during a Zoom call.
“AI is mostly found in neural networks and proprietary software. It looks for patterns and context, so we understand the context of a document – not just a Word document, it can be a tweet , a speech or a message – and we convert the document into a set of vectors. These are 300-dimensional vectors in mathematical space. If you try to convert the vector into 3D, a document becomes a ball dots in space and you are looking for similarities and relevance.
“Basically, we’re doing mapping, we’re using mathematical techniques to look at that data, and a lot of information surfaces that you wouldn’t get if you rely on keywords, which is what Google uses.
“Our customers have very, very complex data – for example with scientists or product engineers – and keyword research is not enough. We understand the entire document, plus the software is completely language independent, everything goes in the math space so you can query an Italian document and get the result in English.
“In the old world, you had to do keyword research in different languages - we don’t do translation, we use trained neural networks in multiple sets of languages.”
Jon says he “can’t take credit” even though he is one of four company founders who is a finalist in two categories of the Cambridge Independent Science and Technology Prize 2022 – AI Company of the Year and Start-up of the Year.
“We’re pretty early at this point,” Jon says of this latest round of investing. “Usually you hope to do it in three months. It will go almost entirely to product development – 70% to research and 30% to sales and marketing.
The potential is vast.
“We can track the metaverse even before it has a name,” says Jon. “There was no such term five years ago, but we understand concepts and context without using keywords, so we can tell a client ‘there is a bunch of stuff going on here, and here are some terms used in the band’.”
Clients, including Mott MacDonald, were suitably impressed.
“We expect customers to be there forever, even though they usually sign a three-year contract.
“We are a start-up company, and many large companies are joining us.”