In 2013, behavioural economist Dan Ariely likened big data to teenage sex: “Everyone talks about it, nobody really knows how to do it, everyone thinks everyone else is doing it, so everyone claims they are doing it,” he said.
Four years later, Michael Connaughton, Oracle EMEA’s head of analytics and big data, argued the analogy should be “put to bed”, as the “spotty adolescent” had grown up and organisations were already reaping the benefits of big data and better analytics.
What’s happened since is the rapid spread of AI across multiple sectors, and while the big tech companies are already well advanced in terms of engaging with its possibilities, there’s now an emerging ecosystem of start-ups harnessing its potential for everything from fintech to medical diagnostics.
Global spending on AI is expected to double to more than $110 billion (€92.6 billion) by 2024, according to the International Data Corporation, and it says organisations will adopt AI not because they want to, but because they will have to if they want to remain competitive in the digital economy.
Certain aspects of healthcare are ripe for AI, and earlier this year Dr Alan Kennedy, an expert in computerised electrocardiography, left his job with Philips Healthcare to set up the Belfast-based medtech company Pulse AI. Its focus is using AI to transform cardiovascular care.
Currently, clinical staff spend hours analysing the results of cardiac monitoring to detect abnormalities. This painstaking process is necessary but slow, and it limits how many patients can be seen and how fast they can be treated. With Kennedy’s solution, the algorithms do the heavy lifting by identifying diagnostically relevant abnormalities at speed. This cuts the analysis time required from hours to minutes.
“We use AI to help clinical staff analyse the signals, and this in turn improves access to and the accuracy of cardiac diagnosis. Our technology supports the human, it does not replace them,” Kennedy says.
“Our product is disruptive as it integrates AI with the latest wearable, mobile and cloud-based technology, and in so doing improves the efficiency of cardiac diagnosis. Most of the current technology does not integrate well with existing clinical workflows, whereas Pulse is a standalone, cost-effective end-to-end solution.
“The idea for Pulse came from a bad personal experience during tests for a suspected cardiovascular issue,” Kennedy adds. “The costs, waiting time and errors I experienced highlighted a major problem I felt could be fixed by augmenting the clinical workflow using AI. Our product will transform how cardiovascular disease is detected globally.”
Kennedy set up Pulse AI in January, and clinical trials are scheduled for December, with the commercial launch to follow in mid-2021. Kennedy sees partnering with a hardware company making monitoring devices as the most likely route to market. By the end of 2020, roughly €110,000 will have been invested in the company, with support from Invest NI and the EU under its Eastern Corridor Medical Engineering Centre initiative, which is focused on improving outcomes for heart disease.
Getting businesses to adopt AI has been a hard sell, but AI start-up Nightingale HQ has developed an online platform to help. Its co-founders, Steph Locke and Ruth Kearney, met while working together on data science bootcamps for the corporate sector at DCU’s Talent Garden.
Locke is a data scientist based in Cardiff. Kearney is based in Dublin and has 20 years’ experience of accelerators, incubators and commercialisation. When they established Nightingale HQ last year their intention was to co-locate in Ireland and the UK. However, Covid has stalled the Irish operation for now. Investment in the business to date is about €160,000.
“Our aim is to help companies overcome the core blockers to successful AI adoption, and rather than doing this individually which would take forever, Steph developed a platform that any company can access to get up to speed,” says Kearney.
“The pandemic has created a valuable change in mindset towards AI, with a much bigger focus now on using it to enhance customer service and increase productivity. Companies of all sizes are becoming more aware that AI can help them scale, innovate and become more agile.”
When the potentially catastrophic impact of Covid-19 on businesses became apparent, Nightingale launched a free initiative called GoSmarter. It’s aimed at all SMEs and designed to help them automate time-consuming tasks quickly. Nightingale provides support and the kit includes tools for sales, invoice processing, social media listening and FAQ chatbots.
Fintech start-up CoalFace Capital is using AI and deep learning to achieve the investment alchemist’s dream: higher returns at lower risk. The company was founded in 2015 and employs 10 people at UCD’s Nova innovation hub. Between founder equity and HPSU support from Enterprise Ireland, investment to date is around €2.5 million.
Initially, CoalFace intended using AI to identify and track a group of consistently top-performing small traders, with a view to developing its own investment fund. However, three years into pursuing this goal, the company pivoted and its focus since then has been on identifying how to extract buy/sell signals from big data to power real-time investment decision-making. The aim is to beat the average performance in any given market by around five to seven percentage points, but crucially at a relatively low risk, as measured by the widely used Sharpe ratio of risk-adjusted returns.
Apart from co-founder and chief executive Declan McEvoy, who comes from a strong capital markets trading background, the team includes graduates with expertise in high-frequency trading, and the company has also drawn on the skills of CeADAR, the national centre for Applied Data Analytics and AI headquartered in UCD, during its development. It has also worked with the Newry-based market analytics software company First Derivatives.
McEvoy says that constantly changing markets require “dynamic” AI to identify situations which, although they might look similar, will never be exactly the same. This differs to the challenges addressed by “static” AI which is used in less volatile spheres such as facial recognition.
“If you combine trading and technology you get a much better result,” McEvoy says. “Some of our ‘secret sauce’ lies in having people with different skill sets address the same problem.”
The way forward for CoalFace is likely through a joint venture, and the company is talking to global asset managers and large insurance companies about a tie-up. All going well, McEvoy is optimistic that CoalFace could eventually create a substantial number of jobs – not least because its IP may have other financial services applications. Insurers, for example, may be able to use it reduce the margin applied to premiums for uncertainty, a saving that could be passed on to the customer through better pricing.