Health data isn’t just helping humans
Unless you have a pet, you’re probably unaware of just how innovative the pet care sector is. Smart collars, cat flaps, feeding bowls and toys are helping owners to monitor the wellbeing of their animals, from cats to cockatoos. Just as preventative healthcare has emerged as a central theme in human healthcare, the same principles are now being applied to pets.
Foundry4 met up with Vet-AI‘s cofounder Paul Hallett, Chief Data Scientist Trevor Hardcastle, and CTO Josh Sephton to find out how the startup is positively changing animal health.
A pioneering platform for pets
According to cofounder Paul Hallett, Vet-AI’s overall aim is to improve animal welfare, the welfare of pets, and to make pet care more accessible… But it’s also about helping those who provide that care. Much like human healthcare professionals, vets often experience difficult working conditions. Through the provision of digital services, Vet-AI wants to relieve the pressure on vets so that they can ultimately offer a higher quality of care. At the same time, this means fewer vet bills for owners, and less discomfort for pets that, generally speaking, don’t enjoy being bundled off to a clinic.
So, how does the platform work? It’s all about the data – by collecting relevant information from veterinary records and from vets themselves, Vet-AI can provide high quality digital care that means that owners can understand and treat their pets from the comfort of their own homes.
“We were very careful from the start to make sure that we are clinically safe and providing the same level of service you would get in clinic with a real human vet,” says CTO Josh Sephton, who previously headed up AI at Push Doctor. “Our research suggests that in quite a few cases, animals don’t need to go and see a vet in the first place because the issue could be treated at home with a shampoo or a different type of food.”
Vet-AI recently launched its first service, a direct to consumer app call Joii. Initially, the app will provide a comprehensive symptom checker for animal ailments.
“We sat a bunch of vets down and asked them all of the questions that they would ask someone if they presented them with an animal in clinic. We’ve structured that into an app in a way that will guide a user through it. There are yes/no questions, severity scores and ratings, mainly to do with dermatology issues to start with like hair loss and skin conditions. They’re quite easy to diagnose for a lay person,” says Sephton.
Eventually, the app will also facilitate video consultations. These consultations could then lead to digital prescriptions, cutting out the need to travel to a vet, pay the fee, and then source the recommended treatments. At the moment, Sephton explains, the Royal College of Veterinary Surgeons doesn’t allow for remote prescriptions.
“We’re currently talking to the Royal College of Veterinary Surgeons about changing their guidelines because, for a human, you can prescribe remotely and have that prescription delivered to your house, without seeing a doctor in person. But, for an animal, you can’t. That’s something we’d like to change over time.”
Vetting artificial intelligence
Vet-AI relies on data to build remote services that can safely and reliably help owners to take care of their pets, but it’s AI that turns this data into value.
“Basically we’re using Bayesian networks to calculate the probability of an illness being present given certain input information,” says Hardcastle. “Essentially it’s trying to emulate the way a vet thinks. When a vet is in practice, they’ll examine an animal, and find out certain bits of information. So they could have a seven year old labrador that has been treated with a certain product, and is showing signs of alopecia. A vet collects all of this information and assesses it to find what they think the most likely illness is. A Bayesian network does the same, but mathematically, using probability theory.”
Rather than replacing vets, though, the network is best positioned as a tool for vets to use. And, Hardcastle adds, building an AI that can replace a human professional is not yet possible. This is why fears surrounding AI are currently, in his view, misplaced.
“It’s self contained and it’s just a piece of code on a computer,” he says. “It’s not going to turn into a Terminator scenario – it’s just not capable of it.”
The cat’s out of the bag…
After a period of stealthy growth, Vet-AI will begin to offer a wider range of services that will support the Joii app. Instead of simply recommending a certain food type or shampoo, Joii will connect with suppliers so that users can instantly order the products they need. Eventually, Vet-AI hopes to build a strong customer base in the UK, and then expand overseas.
Given the hierarchical nature of the veterinary profession, it won’t be a walk in the park for the ambitious startup. The dominance of a small handful of corporations makes it a challenging market for startups, let alone those who want to disrupt the established order.
“The corporates might not particularly like what we are up to, because they’ll see it as disrupting the industry,” admits Hallett. “They’re worried they will lose money, but actually we believe the opposite. Our business model may appear to conflict with that strategy, but we believe we will drive more work into practice.”
Another problem, adds Hardcastle, is that over half of the members of the Royal College of Veterinary Surgeons are nestled in the pockets of big businesses.
“The Royal College of Veterinary Surgeons is meant to be an impartial organisation that oversees the standards of the profession, but 11 out of the 20 members of the RCVS’s council are associated with corporations or are practice owners. They have a financial interest in the industry and maintaining the status quo.”
This could be part of the reason why change has been so slow. Even so, Hallett, Hardcastle and Sephton are positive that Vet-AI can work collaboratively with the Society to demonstrate evidence for change.
Following the successful launch of their app, Vet-AI has proven that the same technologies used in human preventative healthcare can be equally beneficial for animals. Through this, and also by easing the burden on owners and veterinary staff, the startup has built a purposeful company that addresses the challenges of animal wellbeing on numerous different levels.
However, despite the various issues associated with traditional veterinary services, it may still be difficult to coax pet owners away from their habitual vet visits. And, as the market for preventative animal healthcare grows, Vet-AI is likely to face competition from both legacy corporations and new businesses. Nonetheless, the Vet-AI team is confident that the sheer amount of data supporting their platform makes them an entirely different breed.
“Our AI is going to gain insight that has never been attained before,” says Hardcastle. “We are building the biggest and most reliable global pet healthcare database on the planet, period. That makes us different because we will have more data than any individual practise. It’s about the quality and consistency of the data, so we can start creating preventative intervention strategies.”