From animal health to making a new toothpaste, artificial intelligence is helping to solve all kinds of problems.
Artificial intelligence is amazing. Algorithms can take a set of data, learn from it and then use what they’ve learnt to spot patterns, predict future outcomes or even identify objects. While this might sound great on paper, it can be hard to imagine the variety of applications a technology like this can have. These seven Leeds businesses prove the potential of machine learning really does have no limits.
Scaled Insights uses behavioural AI to analyse people’s speech, then work out how to communicate with them in the most effective way. Their behavioural AI tool analyses a patient’s language and infers personality attributes, thinking styles, values and sentiment. As a result, it can differentiate between patients who’d usually be grouped together.
For example, there are two gentlemen, both born in 1948, around three weeks apart. Both men appear identical from a mathematical and data science perspective, since they love dogs, have both been married twice, are in the top 1% of high net worth individuals. However, one is Prince Charles, and the other is the Prince of Darkness, Ozzy Osbourne.
A human being would intuitively understand that Ozzy and Charles aren’t all that similar, but the big data model would put them together. If you’re hoping to influence one of them to take their medicine, to stop smoking, eat healthier, exercise more, buy a car or renew a mortgage, it’d probably make sense to use a different approach to communicate with them to achieve the desired outcome.
Vet-AI is working to make pets healthier by preventing worried owners from googling symptoms when their animals are unwell and bringing real advice to their homes.
The idea was born when co-founder, Paul Hallett, was in a waiting room of a veterinary practice. He overheard a distressed pet owner arguing about the price for their treatment, and the argument was making Paul’s dog so nervous it started shaking. Paul wondered whether the price of care had to be this expensive, and if a pet could be treated remotely to avoid becoming so upset.
The company began in May 2019 with the launch of Joii, an app for people to diagnose and receive treatment for their pets at home. Now Vet-AI is working on artificial intelligence, due to be launched in 2021, that uses data gathered by the app to improve diagnoses. The machine learning models learn the subtle patterns in its vast database. It works out the reasons behind these patterns, learning from images, videos and patient data. The goal is to develop tools that will be used by vets to gain unprecedented insight into animal health.
At the Search Laboratory agency, they’ve used AI to improve the performance of their clients’ digital marketing campaigns.
They’ve used AI in a few ways, including to measure how relevant webpage content is to the key search terms being targeted and to predict the value of an online lead. They’ve also used it to identify the value of website users. Take high-value purchases for example. A typical user visits a website several times before submitting an enquiry, and not all visitors lead to sales, so it’s important to prioritise clients’ remarketing budgets towards the most valuable visitors with the highest chance of converting.
The model scores all sessions based on behaviour, events and page views for each user in both current and previous sessions. They use machine learning to identify the people most interested in the product, then designate them as suitable audiences for retargeting. As a result, they can create tiers of website users based on how likely they are to buy a product.
Security company Videcon has developed deep learning technology to monitor the number of people inside a shop and it’s been used in Aldi stores across the UK during the coronavirus pandemic.
The technology, called Occupi, gives fully automated control over the occupancy level of a building, using AI to count the number of people entering and leaving to maintain social distancing. “The system integrates with automatic door openers to open and shut the door depending on the occupancy level,” Managing Director of Videcon, Matt Rushall, told us. “It includes visual or audio notifications to state when the store is safe to enter.”
At Videcon, they believe that AI has the power to revolutionise the security industry, primarily by reducing false alarms. It gives them the ability to reliably identify moving objects, particularly people and vehicles, which means they can create technology that accurately identifies issues, giving significantly more confidence that the alarms you’re receiving are for events that matter.
FactorAI has written an AI algorithm that tells manufacturers how to develop their products whilst cutting costs and environmental footprint.
The company focusses on formulated products, which means anything that combines more than one ingredient. This could be anything from toothpaste to shampoo or mayonnaise. Changing any single ingredient in a product like this could have a big impact on, for example, an environmental footprint, but it could also change the product entirely. In the past, changes like this have involved a long-winded process of trial and error.
FactorAI speeds it up by retrofitting sensors inside the machines that make the products. “These sensors relay real-time manufacturing data to the workers on the factory floor and to the R&D engineers,” Co-founder Jujar Panesar told us. “Then they’re able to make real-time data-driven decisions eliminating the need for costly and time-consuming trial and error.” This includes the quality of ingredients used and their effect on the final product, manufacturing times and the quantities of ingredients.
CheckLate is a FinTech business that uses AI to predict how good a company will be over the next six months at paying invoices.
It was first launched in 2019 to provide a different form of credit checking facility for companies. Instead of focusing on whether or not a business can afford to pay an invoice, the technology takes data from Companies House along with real supplier reviews. It scores the company on how good they are at actually paying on time, with a simple “Client Score” out of 100. Since launch, they’ve tracked over 4.2 million company records and the technology has been used by thousands of small businesses, helping them decide whether to take on new clients.
This data is then fed into an AI programme called PredictLate, which learns not only how good a company is at paying their invoices, but also matches it against others in the same industry and location. “We then use this algorithm to predict how good the company may be over the next 6 months,” Steve Pritchard, Founder of CheckLate, told us. “This helps companies to understand how much risk there may be in the future. It provides an additional level of data which, in these most challenging of times, will help to minimise financial risk for many.”
Well Good is a tech start-up developing an AI-driven platform to prevent the onset of mental health issues.
“We use personal data to understand someone’s current mental state, then monitor this using AI to predict if they will have a mental health concern,” Well Good’s co-founder, Adam McNichol, told us. The system uses signposts, delivered digitally, to nudge those people towards behaviours that will have a positive impact. The idea here is that they can help prevent mental health issues, but also, stop them from becoming worse in people who are already suffering.
The initial target market is businesses, as the tech will allow them to take a ‘mental health temperature’ of their employees. It uses a chatbot, which communicates with users, learning from the responses they give. “What sets us apart is that we acknowledge that everyone’s personal experience of mental health is different,” Adam explained. “AI allows us to recognise this, that people need help that is truly personalised and not a one size fits all solution.”