How Notable AI Eliminates the Administrative Burden in US Healthcare

In 2017, when Pranay Kapadia was trying to get his healthcare startup off the ground, he sometimes spent entire days cold-calling hospitals and clinics across the Bay Area, trying to get someone — anyone — to meet with him.

As an outsider with no healthcare experience, few people wanted to, even though he promised he could help them cut costs. Those who took his call offered the same feedback: This is healthcare, costs just keep going up. No one’s been able to do this. You won’t be able to change anything

The pessimism was understandable. Kapadia wanted to use technology to automate many of the administrative tasks that burden the US healthcare system — repetitive busywork that accounts for anywhere from 15–30% of all healthcare spending, causes profound levels of frustration for both patients and staff, and takes time and resources away from actual healthcare. Many before him had tried and failed. “Coming from outside in this industry is virtually impossible,” Kapadia says. “It was always, not able, not able, not able. So we decided to use that for our company name.”  

Today, Notable, which automates more than one million repetitive workflows every day, is doing much of what Kapadia and his two co-founders set out to do: working with major healthcare providers like Intermountain Health, CommonSpirit Health, and Sanford Health to automate functions like appointment scheduling, referrals, insurance authorizations, and care coordination.  The Notable platform is live at more than 10,000 sites of care across the United States and has served more than 32 million cumulative patients.

“Fast-forward seven years and we’re one phone call away from pretty much any health system in the country,” marvels Kapadia. 

This success didn’t come quickly or easily. Kapadia’s journey at Notable was punctuated by multiple failures, dramatic about-faces, and cold sweats in the middle of the night. The only thing that got him and his team through, he says, was “sheer grit” — a stubborn refusal to give up and an unwavering belief in the ability of technology to change the world. 

The superpower of software

Kapadia’s fascination with computing started early, while growing up in Mumbai, India. He was exposed to computers thanks to a Sinclair ZX80—one of the first home computers, which plugged into a black-and-white TV and whose software would load through cassette tapes—that his mom had access to from her job as a tutor. To the 7-year-old Kapadia, it seemed like magic. “I absolutely fell in love with the notion of being able to dream up something and then put it into action,” he says. “It was like having this superpower.” 

After getting a computer science degree at the University of Wisconsin-Madison, Kapadia enrolled in the Wisconsin School of Business, where he made the kind of fearless and foolhardy move that would come to define his career. At a lunch hosted by Intuit, he started a conversation with a senior engineering leader at the organization. Kapadia made an impassioned argument for why Quickbase, which the company owned at the time and characterized as a database, was not, in fact, a database. “I was dorking out, but if I strongly believe in something, I’m happy to voice it and go after it,” Kapadia says. Although he had shown up at the event largely for the free pizza, Intuit flew him out to their offices in Mountain View several weeks later and offered him a job. 

Kapadia stayed at Intuit for seven years, working on ways to make QuickBooks easier to use and building Mint, the company’s now defunct personal finance management site. Then one day he saw an opportunity to do something completely new and more than a little difficult. In 2013, he and his wife applied for a mortgage to buy their first home. It was, he says, a “terrible, no good, very bad” experience. After weeks of emailing reams of paperwork and interminable phone calls, they were rejected and had to start the process over again with another lender. “My wife commented that it was harder to get a mortgage than deliver our first child, and that involved 30 hours of labor,” he says.

The experience prompted Kapadia to consider how software could make the process of getting a mortgage less awful. He joined the fintech company Blend as part of the core founding team building the Blend platform. The company, which provides banks and other lenders with software to automate, personalize and streamline the mortgage process, is now publicly traded. 

Into the healthcare unknown

The moment at which Kapadia decided to tackle an even bigger challenge came at a family dinner over the holidays where the conversation turned to the painful, time-consuming aspects of practicing medicine. Six of the assembled family members were doctors. “They were all asking, ‘Why is it so archaic and manual? Why do we have to collect so much data?’” Kapadia recalls. His wife, a psychiatrist, said she felt like “the highest paid data collector in the world.” 

Again, Kapadia considered how software, particularly AI-based programs, could help. The introduction of electronic health records was supposed to have eased the administrative burden, but costs continued to soar. A single healthcare encounter for a patient still required administrative tasks from dozens of people: to do the scheduling, insurance verification, chart review, billing, payment processing, and more. And there simply aren’t enough of these people to go around. Nationally, two million jobs are currently open in healthcare administration, which has a 27 percent attrition rate. “Nobody wants to work these jobs,” says Kapadia. “It’s a completely broken system.”   

With two of his Blend colleagues — Adam Ting, now Notable’s chief product officer, and Justin White, Notable’s chief technology officer — Kapadia set up shop in his Burlingame garage and rented a small office on the Stanford campus. They approached the effort with “child-like curiosity,” seeking to learn all the details of the problems providers and healthcare companies face. The product they carefully developed was focused on addressing provider burnout. Although similar to products Notable now provides, it didn’t sell. The technology worked, but the go-to-market approach didn’t. Instead of selling the ways automation could help doctors and healthcare companies, Notable highlighted improvements to the patient experience. “In hindsight, this was the most obvious mistake,” Kapadia says.

Seize the moment

After the stumble, Notable pivoted to focus on doctors. Engineering and product teams developed an automatic speech recognition tool for the Apple Watch that doctors could use to dictate their patient notes so that records could be automatically uploaded into a patient’s chart. The technology was effective — several years ago, Apple CEO Tim Cook featured it in a presentation — but, once again, it struggled to gain traction. Notable had to sell both healthcare administrators and doctors on the product, with most companies requiring doctors to pay for it.   

“There were a number of times where we were like, ‘We’re done. I don’t know how to sell into healthcare. I’m the worst salesperson that there could ever be,” says Kapadia. 

Then Covid provided Notable with a significant inflection point. In March 2020, after hearing about the chaos and suffering one of his physician family members was seeing in Chicago hospitals, Kapadia huddled with his co-founders, then called an all-hands meeting. He announced that everything everybody at the company was working on would immediately stop. All efforts would now go into helping the healthcare industry meet this unprecedented moment. It was the right thing to do, but was also where market momentum was strongest. “It was one of the hardest things, to tell everyone to stop working on what they had put their blood, sweat, and tears into,” Kapadia says.  

Within 48 hours, engineers had created software to identify patients most at risk for Covid, since tests were not yet widely available. The company also developed a solution that tracked and predicted the spread of the disease, so healthcare companies could better allocate resources like masks and PPE. Later in the year, when vaccines became available, Notable built a platform for automated vaccine appointment scheduling. Using the platform, a joint venture between North Kansas City Hospital and Meritas Health resulted in the scheduling of 1,600 appointments within the first minute of availability, without the hiring of additional staff. Notable also did vaccine scheduling for employees of the medical record company Cerner (now Oracle).

Greylock partner Jerry Chen, who led Greylock’s seed investment in Notable in 2017, says he wasn’t surprised by the pivot. “A startup’s ability to be nimble and seize the moment can make or break their trajectory. Notable’s speed and Pranay’s execution has proven to be of their greatest strengths,” he says. 

Back to the beginning

With wind in its sails and valuable new customers and partners, Notable has spent the past several years working to get back to where it started: automating all the mundane, day-to-day administrative tasks that suck time and money out of the healthcare system. This time, advanced AI technologies like large language models feature prominently. Kapadia says his vision for AI in healthcare is digital assistants either automating or enhancing every administrative function. Notable is already doing some of this for its customers. The company’s scheduling assistant, for instance, allows patients to make appointments automatically using a provider’s website or app, after its authorization assistant has confirmed insurance coverage. Its care gap closure assistant instantly analyzes a patient’s electronic health record to identify when someone might be overdue for a mammogram or hasn’t renewed a prescription, allowing a human care coordinator to review thousands of patient charts per day, instead of dozens.

“Imagine not having to make a phone call ever again in the healthcare system, not for scheduling appointments, verifying insurance, making referrals, or collecting payments. That’s what we’re working on,” says Kapadia.

In the future, Kapadia sees value in AI doing even more to proactively identify next steps in a patient’s care, thus helping catch potential health issues early and improve outcomes. This could mean identifying and scheduling a patient’s follow-up appointment before they leave the clinic, or enrolling a patient in a particular care program, with all of the associated work to make this happen handled automatically.

“We’re not trying to replace the physician’s autonomy and decision-making. There’s nothing that can replace that human connection and empathy,” says Kapadia. “We’re trying to make it even better.”