Samir Manjure never believed he would one day start his own company. “I was at Microsoft for 17 years, held various positions, grew up the ranks, building AI and machine learning solutions for enterprises,” says the co-founder and CEO of Seattle-based healthcare AI analytics company, KenSci. “I was at a stage where I wanted to do something that has meaningful impact on society at large. That’s what got me to look specifically at education and healthcare.”

Manjure conveyed his thoughts to childhood friend, Ankur Teredesai, Professor of Computer Science & Systems at the University of Washington Tacoma. For the past decade, Teredesai’s research work has focused on creating assistive AI in the healthcare setting and utilizating AI in the prediction of chronic conditions including heart failure and diabetes. “I was fortunate to meet Teredesai,” Manjure says. “We agreed on so many things, including how healthcare is so underserved and how we can leverage data science in a way that can benefit”.

KenSci was born in May 2015 (Ken refers to knowledge, Sci means data science) with Teredesai appointed as the CTO. “Our goal is to solve real, clinical challenges, not to show how sexy machine learning can be,” says Manjure. He had noted that a lot of the healthcare information being collected from patients was not fully utilized to make informed decisions. This is where KenSci’s opportunity arose. It harvests data from electronic health records, insurance claims and other sources to come up with specific applied insights to facilitate the generation of cost-effective interventions.

“Our mission is to ‘fight death with data science”, but we don’t just tell healthcare providers, ‘Here are the patients with the highest risks for readmission’,” Manjure explains. “Instead, we say ‘Here are the five factors you should consider when you meet a high-risk patient stratified by our software’. We arrive at these suggestions after breaking down the health data into components based on medical history, vital signs, lab results, psycho-social metrics, and so on. In a way, this made AI ‘explainable’. At the same time, we are also establishing partnerships with healthcare providers and their IT force.”

Manjure believes AI works differently in healthcare when compared to other industries. The technology will not only be assistive, but it also has to fit seamlessly into the workflow. “Healthcare providers do not have time to click, read and understand one more dashboard or analytic tool in their practice. Our solution needs to be consumed readily without friction.”

Right now, KenSci focuses on three main verticals: care management, operation and cost. The company builds disease progression models to predict patients with a higher propensity to develop chronic illnesses or evolve into adversity. It also assesses the whole care continuum to forecast all kinds of operational processes to ensure sufficient resources are available. “A few years ago, the common question was, ‘Why do we need AI in healthcare?’ Today, the conversation has shifted to ‘How can AI help us?’” explains Manjure. “So, we are here to prove that algorithms can impact patients’ care journey and save lives.”

Nonetheless, after harnessing $22 million Series B funding in 2019, Manjure believes that for KenSci, it’s still early days. “If you think about your experience with a music streaming or e-commerce platform, they can make predictions on what we are likely to listen to or buy next using very little information,” he says. “While in healthcare, we have decades of data so we should be able to do more. Yet, our utilization only began in the last decade.

“That’s why KenSci chose to focus on small things. There are two kinds of AI companies out there. The first uses data sciences to fulfill bigger claims like curing cancer. The second is very focused on the basics. KenSci belongs to the second kind. We want to fulfill minute tasks that are applicable and can significantly impact many patients.”