AI Startup Olive Specializes in RPA for Healthcare
About four years ago, we first came across the concept of robotic process automation (RPA), which originally sounded to us like some marketing hack had repackaged software automation around the hype in artificial intelligence. A couple of years later, the RPA market suddenly exploded. Gartner estimated revenue grew by 63.1% in 2018 to $846 million, making it the fastest-growing segment of the global enterprise software market. We started finding all sorts of startups using AI for automating back-office jobs like invoicing
and gossiping. These include well-financed unicorns such as Automation Anywhere and UiPath, which is pushing the idea of hyperautomation to digitize every possible office job. Another RPA startup that just joined the unicorn club this month, an Ohio outfit called Olive, specializes in using AI to automate repetitive, high-volume tasks in the healthcare industry.
AI in Healthcare Raises Record Funding
Founded in 2012, the Midwestern startup has amassed $448 million, with $385 million coming in the last nine months during you know what. The $225.5 million round this month pushed the company’s valuation to $1.5 billion. Investors included well-known venture capital firms like Sequoia and General Catalyst, as well as Google (GOOGL). The size of the round is impressive, even in a year that has already seen a record-breaking $4.7 billion in funding to AI healthcare startups in 2020. In the third quarter alone, companies took in more than $2.1 billion across about 121 deals, according to research firm CB Insights.
In fact, Olive ranked No. 5 among the top AI healthcare deals last quarter:
In a recent but separate report on AI trends in healthcare, the big brains at CB Insights said AI and RPA will help drive efficiencies in an industry where administrative costs eat away at value-based care. The report specifically mentions Olive, and notes that while some of the hype around RPA technology has plateaued, hospitals are only now buying into this type of automation.
“This could be attributed to the fact that the majority of RPA vendors today are general-purpose solution providers, catering to a wide range of industries. Few startups are specifically designing platforms that work seamlessly with the technical and regulatory bottlenecks unique to the healthcare sector.”CB Insights
We threw a bunch of stats at you in a previous piece we wrote about AI healthcare startups digitizing hospital operations in which we briefly profiled Olive. But here are a few other interesting numbers to consider when you wonder why so much money is being invested in algorithms that auto-complete medical forms at scale: The United States spends 34% of its healthcare dollars on administrative costs. That works out to about $2,500 per person, according to a study that showed just how much less the Canadians are spending on healthcare overhead.
That kind of money means you’re paying a lot of people to push paperwork. In the jobs category Medical Records and Health Information Technicians alone, the Bureau of Labor Statistics estimates that about 340,000 people are gainfully employed at a median salary of more than $42,000. And that’s just one labor category in the vast healthcare bureaucracy that buoys perhaps the most complex, lucrative, and absurd marketplace on the planet. What if there was an AI workforce that could read digital medical records, scan visual or textual imagery, and even figure out how to play the insurance game more
ruthlessly efficiently? That’s the business proposition from Olive that has reportedly earned it 600 clients and counting among healthcare providers.
How AI Supports Olive’s RPA Platform
The company actually uses AI to automate and analyze workflow in several ways. There’s the basic algorithms that can do the repetitive stuff like copying and pasting data, inputting and correcting data, and solving complex calculations. That’s RPA 1.0. Olive also uses computer vision to capture metadata from screens and visual interfaces, connecting that information to the task at hand, as well as for data to train the neural networks that look for patterns to find trends (actionable data, in the parlance). How much data? Over the last three years or so, Olive has ingested more than 100 terabytes of data. For comparison: The Hubble Telescope, in business for three decades, has collected more than 150 terabytes of images and data and adds approximately 18 gigabytes of data weekly. And that’s the whole friggin’ universe.
Big Data, Big Payoff
All that data feeds machine learning and deep learning systems that provide different kinds of insights. For instance, machine learning finds trends to determine which procedures and payers have the highest denial rates and adjusts the provider’s strategy to reduce or prevent those denials in the future. In an example of the platform’s deep-learning capability, Olive could scan medical forms to understand what type of documents and data it is viewing and associate words or images with billing charges and outcomes. Over time, all that delicious data will improve the AI’s ability to actually predict diagnoses and the future costs of all that hair care.
In a case study on a national healthcare client, Olive ingested a historical dataset that included 30,000 PDFs containing more than 200,000 different types of patient documents. Olive went beyond simply shuffling paperwork, eventually training itself to classify unfamiliar documents, which helped automate the process for determining patients’ insurance eligibility, as well as updating eligibility status in the organization’s records. That significantly sped up the insurance verification process.
The Verata Acquisition
Olive bought its own kind of insurance when it acquired its first company during the same week that it raised the company’s latest mega-round. Olive reportedly spent $120 million to absorb another Midwestern AI healthcare startup in Verata Health. Founded in 2017, Verata raised an undisclosed Series A last year, so it’s difficult to say if Olive paid a premium or got a bargain with the acquisition. However, the four previous investors demonstrate that Verata had some traction: dividend-growth stock darling 3M (MMM); the venture arm of BlueCross BlueShield; the well-respected Silicon Valley VC firm Bessemer Venture Partners; and LRVHealth, a VC firm made up of investors from provider, payer, and vendor organizations, comprising a network that touches one in three healthcare consumers across the United States.
The problem that Verata tackles with AI involves what’s called prior authorization, which is when your doctor or healthcare provider
makes a deal with the devil seeks approval from an insurance company before delivering a specific service, such as life-saving care. While prior authorization costs account for just 2% of overall medical industry admin spending, the process is the most costly and time-consuming, according to a report from the Council for Affordable Quality Healthcare (CAQH). The CAQH says the medical industry could save $454 million annually by transitioning to fully electronic transactions:
Verata’s solution, Frictionless Prior Authorization, is for both providers and payers. The platform connects to electronic health record (EHR) systems and automatically initiates prior authorizations, retrieves payer rules, and helps identify and submit clinical documentation from the EHR. On the payer side, Verata enables point-of-care authorizations for providers and patients. Olive claims the combined AI-powered prior authorization solution between the two startups will reduce write-offs by over 40% and cut turnaround time for prior authorizations by up to 80%.
New AI Healthcare Services from Olive
Olive was already expanding its services before the Verata acquisition. Last month, it introduced something called Olive Helps, which delivers real-time targeted information to employees so they can work better and faster. How? Frankly, it’s not entirely clear to us, but the new AI platform supposedly understands and optimizes human behavior in the form of what’s called a “cybernetic loop,” which can be created and accessed in Olive’s Loop Library, a sort of app store. These Loops bring together the intelligence of healthcare providers and technology communities across the country, according to a company press release on Olive Helps.
It sounds like there’s potentially a lot of work involved in integrating all of these hyperautomation healthcare solutions. Don’t worry, because Olive has that covered with AlphaSite, which is a co-located team of the company’s own engineers and project managers who work to integrate AI throughout the organization. Each AlphaSite comes with an onsite AlphaWall displaying Apertures, Olive’s real-time data visualization tool. The idea is that everyone in the organization can watch how much money Olive is saving it 24/7. The company is so confident in its tech that each AlphaSite comes with a 5X ROI guarantee through a performance-based contracting model.
Olive has deployed 11 AlphaSites so far. At Tufts Medical Center, for example, Olive went to work on helping with COVID-19 testing operations. The AI workforce went live in two months and almost immediately increased testing rates by 7.5 times, by cutting the amount of time spent on the testing process by 86%. Another AlphaSite at Yale New Haven Health has two live automations, and the team identified more than 20 other processes, with potential savings of $17 million or more.
Perhaps the biggest story in healthcare this year – aside from you know what – involved Google’s DeepMind Lab. It created an AI system that can predict the complex shapes of proteins based on genetic sequences, with huge implications for drug development. While yet another advancement taking us closer to artificial general intelligence, the Nobel Prize-worthy breakthrough was only possible by creating an AI specifically tuned to the task. In other words, to solve challenges in healthcare, you need algorithms and neural networks designed and trained for those data. Olive could be the right company at the right time for grabbing market share before other specialists show up to the party.
If and when a company like Olive will be open to retail investors is impossible to say. However, those bonkers for bots in the back office can sink money into companies like Blue Prism (PRSM.L) and Pega (PEGA).
Want to know which stocks we hold for exposure to RPA? Check out the Nanalyze Disruptive Tech Portfolio Report which covers dozens of stocks including our own real-money portfolio.