OKRA Technologies founder Dr Loubna Bouarfa has worked in the field of AI since 2007, when she began her PhD at Delft University in the Netherlands. Here, she focused on using machine learning to predict surgical workflow and detect anomalies in real time, and found that artificial intelligence (AI) provided a strong framework for evidence-based medicine. By inputting data to an AI, she could give herself and her colleagues the opportunity to improve people’s healthcare outcomes in the real world, without conducting a lengthy clinical trial or healthcare study to reach a conclusion.
This led her to develop the OKRA platform, an AI platform which allows healthcare professionals to access a combined view of all of their data. It uses this to provide them with insight into how best to deploy their product.
The platform combines data from different sources and learns how patients respond to treatment, figuring out which populations are most in need of certain medications and how to reach them.
“We have a highly secured AI platform with gold-standard measures to secure our accessible data, and all our patient data is anonymised and de-identifiable at point of source,” says Bouarfa. “In terms of data access, we integrate to our AI platform three types of data: open health and environmental data from different countries, our client’s proprietary data and third-party data obtained through licensing and partnerships.”
OKRA’s platform is both figurative and practical
This gives OKRA users the power to boost business by getting the right treatment to the right places at the right time.
OKRA operates on a figurative level, using environmental and health data to make predictions and assessments about in-need populations. The environmental data, for example, can predict a when a heatwave will take place in a certain country, meaning a pharmaceutical company now knows it should target its solutions for illnesses which thrive in heat towards that particular location.
“Pharmaceutical companies can ask OKRA questions about the disease areas they specialise in,” says Bouarfa. “Where’s the population in need? What are the statistics? How many patients are will be in need of this treatment for up to 12 months into the future in any basic region, and why that’s the case? We combine environmental factors with health factors and give answers based on that.”
OKRA can also aid pharma companies by supporting them with healthcare provider outreach.
“Imagine you’re a pharmaceutical company and you want to decide where you need to supply a new treatment. You can use OKRA to look into areas where doctors are not prescribing this treatment and still prescribing an older medication,” says Bouarfa. “We flag them up and provide our clients with insight: how to reach out to doctors, what messaging to use, and what clinical based evidence about the region is available. They can then help patients reach the new treatment.”
OKRA also has a system in place which can be used to recruit patients for probationary drug testing. The algorithm uses clinical trial data to learn and identify which patients are appropriate candidates for a new drug.
“If a man has some type of cancer and we see he’s not responding well to chemotherapy, our system can flag him out and say ‘he should take part in this clinical study, his chance to get there is much higher because he looks like the patients who have responded better in the clinical trial’,” says Bouarfa. “Our goal is finding out how we can provide the right treatment to the right patient, and how we can maximise the speed and precision of doing do.”
Advising EU on the future of AI
The profile of both OKRA Technologies and Bouarfa has grown dramatically since its inception, with the entrepreneur featured in Forbes magazine as one of Europe’s Top 50 Women in Tech 2018. That year, she closed a £4.2m round of Series A funding for the company and was appointed to the European Commission High-Level Expert Group on Artificial Intelligence (AI HLEG).
AI HLEG aims to maximise the benefits of AI across Europe, while minimising the risks. Bouarfa’s role is to contribute to the ethical guidelines, focusing on the boundaries which need to be respected within AI. She also looks into bias control and maximising the competitiveness of AI in Europe, focusing on how to drive the European market and make the continent a world-leader in AI.
However, Bouarfa describes her biggest achievement as being the real life outcomes she’s been able to create with her technology.
“As a start-up OKRA brings a lot of innovation, but we can no way bring this innovation without internal champions working in different spaces in the companies we target,” she says. “For me that’s our biggest achievement, that we could build this bridge with them. An amazing team has built and grew OKRA to where it is today and will take us further in the future.”
What’s next for OKRA?
OKRA is currently working with three global pharmaceutical companies, the identities of which Bouarfa cannot reveal at this time, to help them improve the accuracy of their treatments and market outcomes. While the company’s main focus rests on life sciences, the OKRA platform has even been used to help foster children find stable homes.
Bouarfa says OKRA Technologies is strengthened by her position as a female entrepreneur of mixed race heritage leading an AI company.
“At OKRA we like diversity. Diversity brings different perspectives to the table and continuously challenges our biases, mind-sets and ways of thinking,” she says. “This is exactly what applies to our industry, how can we reach out to the patient that is the outlier in the current healthcare system? So I would say to all start-ups: be diverse, create a diverse scene, so you can reduce bias and achieve the right outcome you’re looking for.”
Bouarfa’s journey into tech has been inspired by figures who have taken uncertain paths to pursue their dreams. She lists Arab migration poetry by authors such as Khalil Gibran and Ilya abu Mad, the story of Dame Stephanie “Steve” Shirley and Siemens chairman Jim Haagman Snabe’s Dreams and Details as sources which have encouraged her to drive innovation further.
And so she should. Machine learning has now surpassed humans in predicting heart attack and death in cardiac patients, can use vocal biomarker analysis to diagnose post-traumatic stress disorder and even detect early warning signs of dementia. When AI functions so highly on a patient-care level it makes sense that it ought to be generalised to an analytics level.
“This is the fourth industrial revolution,” Bouarfa says. “AI is about learning from our humanity and accepting that we don’t know all the facts. AI will transition healthcare into a path of proactive industry, from the current diagnose and treat model to a predict and prevent model, and from a volume-based industry to a value based industry.”