Outbound AI has a prescription for one of the biggest headaches of the medical system: interacting with insurance companies and other payers to process claims.
The Seattle startup emerged Tuesday with a software product, PayerVA Console, to help automate claims processing. The company, which spun out of Madrona Venture Labs with $7 million in seed funding more than a year ago, also plans to take on other rote tasks in healthcare.
Outbound AI’s 27-member team is peppered with veterans of other conversational AI companies such as Saykara, a Seattle clinical voice assistant startup acquired last year by Nuance Communications, which is now owned by Microsoft.
And while Saykara and Nuance mainly support clinical encounters between patients and providers, Outbound AI takes a different tack, said CEO Stead Burwell.
“We’re radically focused on all those conversations that exist pre- and post-encounter,” said Burwell, who was previously CEO of Seattle healthcare data startup MultiScale, which was acquired by Providence health tech spinout Tegria.
PayerVA Console takes some of the tedium out of calls to payers, like staying on hold and culling claims data. The tool senses when it’s time for a human to take over and tells them, and it transcribes calls and extracts the key elements.
“We’re not out-of-the-gate looking to replace human to human calls,” said Burwell of the tool. “What we’re looking to do is expedite, augment and offload some of the burdensome elements between a provider and a payer.”
PayerVA Console can shave off about 40-50% of a full-time phone worker’s time and requires only minimal training, said Burwell. The product will also help customers retain claims specialists by improving their job experience, he said.
Outbound AI is one of more than a dozen spinouts of Madrona Venture Labs, the startup studio associated with Madrona Venture Group that fast-tracks the early-stage process of forming a company and recruiting a leadership team. Spinouts from the lab have been acquired by Nordstrom, Uber, Mailchimp and other companies.
Madrona Venture Labs first approached Burwell for advice about applying conversational AI to healthcare’s “back office,” and later tapped him as co-founder and CEO. Shortly after its founding in March 2021, the startup pulled in its $7 million seed round from Madrona Venture Group, EPIC Ventures, SpringRock Ventures, Ascend VC, Pack Ventures, and strategic and angel investors.
Burwell is joined by three other co-founders: Justin Ith, a former product lead at Madrona Venture Labs, Saykara veteran Kshitij Moghe, and chief technology officer Jonathan Wiggs, who helped build the speech models for Apple’s Siri while at Nuance, and most recently was vice president of engineering and architecture at tax software giant Avalara.
The startup is now onboarding its first customers for PayerVA Console and will market the product to small- and medium-sized healthcare groups, such as independent physician practices and the medical billing companies that support them.
Outbound’s engineering team was helped along by the relatively small vocabulary needed for claims processing, said Wiggs. People typically need about 1,500 words to support a conversation, while the focused conversations in claims processing happen mainly with 200 to 300 words, he said.
“It’s been exciting to see that capability really come to pass, and watching billers in the healthcare space use it and really love it,” said Wiggs of the new tool. Outbound also partnered with Tegria for the design and commercialization of the product.
Outbound joins the growing ranks of companies aiming to use AI to ease the administrative burdens of healthcare.
MajorBoost, a spinout from Seattle’s Allen Institute for AI (AI2), landed $350,000 in pre-seed funding last year to develop a product to ease communication between healthcare providers and payers. Another AI2 spinout, Birch AI, aims to simplify call center work in healthcare, insurance and banking. Providence-linked Advata has a focus on healthcare revenue cycle management, and Paris-based Shift Technology uses AI to detect insurance fraud.
Outbound AI has plans to next help automate authorization of medical services prior to their delivery. It also plans to create software to support care management programs integrated into value-based care models.
With companies on both sides of the provider and payer equation increasingly turning towards automation, does Outbound AI’s new tool bring us closer into a world in which robots talk to robots? Integrating data across systems is a complex problem and speech has a role no matter who — or what — is doing the talking, said Burwell.
“There are going to be conversations in human language that will be more adaptable and easier to execute and operationalize than traditional data models that the data integration world has been working on over the last 20 years,” said Burwell.