KenSci: Using Artificial Intelligence and Machine Learning to Save Lives

While using artificial intelligence and machine learning to optimize health care isn’t a novel concept, KenSci proclaims its mission is more broadly to fight death with data science. By collecting big data, analyzing variables and outcomes, and suggesting cost-effective and successful solutions, KenSci is aiding medical professionals in fending off the grim reaper. Keys to the company’s success is its ability to provide customers proof of a return on investment while improving patient outcomes, as well as providing an easy path for integrating right into the systems a hospital or health facility already uses. 

AI and health care is an active field for investing, as many companies try to tackle outcomes cross optimized with costs, and venture capitalists are paying up. According to Frost & Sullivan, the AI health market is expected to reach $6.6 billion by 2021, growing at a compound annual growth rate of 40 percent. Accenture believes that key clinical health AI applications can potentially create $150 billion in annual savings for the United States healthcare economy by 2026. 

As an example of a recent investment, in January, Health Catalyst raised $100 million at a lofty $1 billion valuation given sales of just $110 million in 2018. This brings its total funding to $392 million. In its recent round, funding was comprised of $85 million in debt and $15 million in equity and is being used to invest in development of its data platform and increase its clinical, financial, and operational consultative teams. It touts sepsis treatment as an example where it has most successfully used AI and patient data to improve outcomes. The key has been to identify and treat correctly it as quickly as possible. This includes rapid diagnosis and making sure the correct treatment is executed. Customers can then track how compliance achieved better outcomes and saved lives.

While most companies are tackling care from the enterprise side, there are even consumer facing companies using AI to help users diagnose themselves. For example, K Health provides an app that asks users questions and then compares them with over a billion AI analyzed patent records to suggest possible causation and even treatment. Founded in 2016, it has already raised $44 million and has partnered with primary care providers to help drive patients to doctors when necessary.

Not all companies using AI in healthcare are venture-backed startups. IBM has a big push in this area via its Watson Health division. The company offers a wide range of solutions for the health care vertical market and has select offerings using AI such as its AI Patient Companion for Hospitals. It is a turnkey solution leveraging IBM Watson and other AI technology to help improve patient outcomes, reduce labor costs, and increase hospital revenue. Labor is the biggest component of a hospital’s costs and nurses are the largest subset. This system, with a Siri type interface, answers questions for patients thus decreasing nursing labor. These questions could be things such as “When am I scheduled to go down for X-rays?” or “What is the lunch menu?” Another AI based offering is Micromedex with Watson that had been used to allow natural language queries of drug databases to help pharmacists get quick answers for drug classes, dosing and administration, medication safety, mechanism of action, pharmacokinetics and drug interactions. 

NVIDIA also has a big healthcare push fueled by imagery in diagnostics that requires high-end graphics processing, its forte. It has introduced its Clara platform that enables developers to build and deploy medical imaging applications. It is the foundation to enable of intelligent instruments and automated healthcare workflows. NVIDIA takes what it learned in the graphics required for gaming and helps customers apply it to the medical imaging industry. Its SDK has been used for processing for medical reconstruction, image processing and rendering, and computational workflows for CT, MRI, and ultrasound images. In November, the company announced it is working with GE Healthcare and Nuance to bring artificial intelligence to their medical imaging solutions. The companies plan to integrate GE installed base of 50,000 machines onto NVIDIA’s platform, and to accelerate the creation and use of deep learning algorithms for its instruments. Nuance plans to bring machine learning to radiologists and data scientists through its AI Marketplace for Diagnostic Imaging, which has been built on NVIDIA’s deep learning platform. It will create algorithms to help the workflow of the radiologists, aiding them to detect and quantify clinical findings more quickly and improve care. Nuance has a large installed base of image-sharing and reporting solutions, and they are now used by 70% of radiologists.

In contrast to the focus on these two large companies, KenSci is more like Health Catalyst. However Health Catalyst has focused on building a data warehouse architecture and uses analytics, while KenSci has focused on AI and machine learning to improve algorithms utilizing the science where its founders excel. Two friends from India, who both earned degrees in computer science from Maharaja Sayajirao University, co-founded KenSci in 2015. KenSci is based in Seattle, Washington and has offices in India and Singapore. Its CEO, Samir Majure, has an MS in Computer Science from Perdue and a Wharton MBA. Before co-founding KenSci, he spent 17 years at Microsoft. In his last position there he managed a portfolio of CRM products and integrated online services for marketing and sales. He also successfully launched new innovative products in Small Business CRM category. In this role he had direct responsibility for managing a multi-discipline product unit of about 75 people, who included developers, program managers, and professionals in Quality Assurance, User Experience and User Assistance. The second co-founder, and the current CTO of KenSci, is Ankur Teredesai. In addition to his role at KenSci, he is a professor at the University of Washington in Seattle where he teaches applied machine learning and is Executive Director of the school’s Center for Data Science. 

KenSci was incubated at University of Washington’s Center for Data Science at University of Washington Tacoma, and designed on the cloud with help from Microsoft’s Azure4Research grant program. The company got its start with grants from DARPA and NASA JPL, and partnerships with leading Health agencies like The Center for Disease Control. The KenSci platform is a set of services that are built on Microsoft Azure ML, SQL and uses Power BI. The services make it easy for machine learning workloads especially in healthcare. It is essentially a scoring platform at the heart of it, but it also has services that make it easy to develop models. It also has bank of service features out of the box with pre-implemented models to be used as a starting point. Today the company employs about 60 people with 90% located in Seattle. It has projects worldwide and boasts projects in the UK, Singapore and India where is seeks to expand further. In fact, during 2018 the company doubled in size.

KenSci had used AI and machine learning to predict the course of a variety of medical issues as shown below, including diabetes, CHF, COPD, CKD, and sepsis. The company’s platform is based on a database of over 150 machine learning model and the algorithms are developed based on over 10 million sets of data. These data sets are clinical data (medical records), financial data (insurance and health care costs) and patient biometric data (for example heart rate). KenSci software reorders and analyzes the patient data to identify patterns that may indicate potential risks and provide predictive information. This technology can then be used to reduce the rate of hospital readmission of patients.

The platform has been used for a variety of custom solutions too. It has:
Improved emergency department staffing and patient experience at Rush University Medical Center
Identified lifetime healthcare costs for 9/11 first responders for the Center for Disease Control (CDC)
Reduced the cost of prescriptions at a system level without sacrificing quality for Beaumont Health in Michigan
And created risk stratification and readmission predictions for military personnel for the Madigan Army Medical Center.

One fascinating project the company took on using CDC data is a look at predicting the opioid crisis using historical data from 1999 to 2015. By taking data from all US counties, the company is able to predict how widespread the problem may become in a few short years as deaths are increasing logarithmically throughout the country with current hotspots in New Mexico and Appalachia.

On Friday, the company announced it had raised its second round of financing. It raised $22 million in Series B funding. The round was led by Polaris Partners, and was joined by investors UL ventures and included additional funding from Ignition Partners, Osage University Partners, and Mindset Ventures. This round brings KenSci’s total funding to $30.5 million. 

Back in January of 2017, the company raised $8.5 million in a Series A round led by Bellevue-based firm Ignition Partners, and included Osage University Partners and Mindset Ventures. The company used the funds to increase its workforce and invest in sales, marketing and customer service to support more customers. 

With its new capital it has quickly moved to strengthen its staff. On the 23rd of January, it announced four strategic hires in engineering. All four have experience in AI, healthcare and cloud solutions for enterprises.

The new hires are:
• Sudarshan Chitre, former Vice President of Engineering at Oracle, who brought with him 20 years of experience building software for enterprise and consumers. In his career, he has had the opportunity to lead teams of exceptional engineers to build v1 cloud platform services for Azure (HDinsight) and AWS (Internet of things). He was also the technical co-founder of Brimbee and held various roles at Amazon. He began his career at Microsoft where he spent 16 years.

• Ryan Brush, former Distinguished Engineer at Cerner, spent the last several years there on data engineering, analysis, and the application of very large healthcare datasets. He created Clara, an open source rules engine, Bunsen, an open source library for large-scale analysis of FHIR datasets, and has spoken at conferences including Strata, Strangeloop, ApacheCon, FHIR DevDays, and others. As an author, Ryan has contributed chapters to the books 97 Things Every Programmer Should Know and Hadoop: The Definitive Guide.

• Neelesh Kamkolkar, former Lead of Product Management at Tableau, brought over 15 years of experience building platforms, solutions, and services. Most recently, he led enterprise product management at Tableau Software, where he built a platform that enabled data analytics to be accessible to everyone in a secure, scalable way. 

• Tim Kellogg, former Senior Software Engineer, at Amazon Web Services, brought with him over a decade of experience. As an experienced full stack software engineer with a proven track record of building performance-critical systems, Tim has a comprehensive knowledge of garbage collection and JIT compilation in .NET that he has used to build scalable applications in C#.

SUBSCRIBE TO ZACKS SMALL CAP RESEARCH to receive our articles and reports emailed directly to you each morning. Please visit our website for additional information on Zacks SCR. 

DISCLOSURE: Zacks SCR has received compensation from the issuer directly or from an investor relations consulting firm, engaged by the issuer, for providing research coverage for a period of no less than one year. Research articles, as seen here, are part of the service Zacks provides and Zacks receives quarterly payments totaling a maximum fee of $30,000 annually for these services. Full Disclaimer HERE.

COMMENTS

WORDPRESS: 0
DISQUS: 0