Apacendo Health co-founders Chong Sun, left, and Jonathan Lin. (Apacendo Health Photo)
Each year, more than 130,000 people reach kidney failure. It’s the most advanced and expensive stage of a disease that affects 37 million Americans, 90% of whom don’t know they have it.
The tools being used to manage those patients, in many cases, haven’t kept pace. At nephrology clinics across the country, critical patient information still arrives by fax. Lab results sit in one system while the treating physician works in another. Staff manually key data into electronic health records one document at a time.
“I can walk through the airport and facial recognition lets me through. I can take a picture of a check to deposit it,” Jonathan Lin, co-founder of Seattle-based Apacendo Health, told GeekWire. “But yet, we still manage this disease state with faxes and Excel spreadsheets. It’s so archaic.”
Lin and co-founder Chong Sun believe AI agents — software that can autonomously navigate interfaces, read documents and take action inside existing healthcare systems — can close that gap. Their startup is building what they describe as an AI-native operating system for nephrology practices: software that works in the background, processing incoming faxes, triaging incoming data, and handling the administrative labor that consumes hours of staff time every day.
A disease hiding in plain sight
The U.S. spends more than $150 billion annually managing the consequences of chronic kidney disease, including over $50 billion on dialysis alone, while the NIH invests $19 per patient in research to understand how to treat and prevent it.
“For the most part, this is a silent disease,” said Dr. Osama Amro, director of nephrology at Swedish Medical Center in Seattle and an advisory board member for the National Kidney Foundation’s Pacific Northwest chapter. “Patients don’t have symptoms of pain in the kidneys or anything that brings them to a physician, except when it’s late.”
A blood draw and a urine sample can detect kidney damage years before symptoms appear. But catching patients at that stage requires a level of coordination and data-sharing that the current healthcare infrastructure isn’t built to support.
“We rely on a very tedious process of reviewing data,” Amro said. “The data, believe it or not, comes in faxes, in multiple locations, despite having an electronic medical record. This is not designed to screen patients or manage them with chronic kidney disease. Many times there is a delay in evaluating patients who need initial evaluation in a timely manner.”
A patient with chronic kidney disease can cost the system about $30,000 per year, with the price rising as the patient reaches end-stage, driven by dialysis and hospitalizations that earlier intervention could have prevented. By that time, a patient without a transplant has less than a 50% chance of surviving five years.
Supporting the back office
Anika Porter has been a nephrology practice administrator for 17 years. As practice administrator at Global Kidney Care in Houston, she oversees the operational side of a clinic where physicians see 20 to 24 patients a day, spending more time with each than providers at many comparable practices. The administrative burden, she said, falls hardest on staff who rarely get attention.
“People are so focused on the physicians,” Porter said. “The back office doesn’t get much support.”
Before introducing Apacendo Health’s technology, her clinic had two people dedicated to managing faxes. It’s a task that can mean hundreds of documents a day, each requiring manual review and data entry.
Dealing with insurance companies adds another layer of friction. Billing codes are standardized, but reimbursements often aren’t. Lin described a practice called downcoding, where insurers pay significantly less than what was billed, without notifying the provider.
“The doctors will perform a service, they will bill for that service, and then insurance companies will pay them much less, and won’t even tell them they’re paying at a discount rate,” Lin said. “Most doctors will never figure out that this is happening until they start reviewing their finances.”
“We’re at their mercy,” Porter said of insurers – not to mention that pay in the specialty has stagnated, and nephrology is among the most susceptible to turnover and budget cuts. In 2023, about 52% of nephrologists in the United States were international medical graduates — a sign of how few Americans pursue the specialty, and how uncertain the field’s future is given ongoing immigration policy.
Updating the playbook
Lin spent years working in the dialysis industry before moving into private equity and venture capital focused on healthcare. Two companies, DaVita and Fresenius, control about 70% of the U.S. dialysis market. What Lin observed was a system organized almost entirely around end-stage disease, with little infrastructure supporting the earlier, more preventable phases.
“A lot of the industry is relying on a high-touch clinical model, where they believe that if you engage with a patient on a very common basis, you can prevent their disease from progressing,” he said. “But the challenge is that it’s very highly manual. We’re basically using the playbook from five to ten years ago and applying it to this problem.”
A machine learning scientist, Sun had no prior experience in healthcare. His entry point was personal: his wife, a Navy veteran who became a VA mental health therapist after retiring from service, was spending up to seven hours a day on paperwork, leaving only three hours for actual patient care. In 2023, Sun built her an app to automatically generate session notes from recorded patient conversations. The VA wouldn’t adopt it, but the experience taught him what it meant to try to change healthcare from the outside.
The two connected through mutual friends and started Apacendo Health in 2025. Lin understood the clinical workflows, payer relationships and political terrain of a fragmented industry. Sun understood how to build software at scale. Their company, now at three employees, works with four nephrology practices across the country. So far, they’ve raised an undisclosed amount from the Science Fair Fund and angels.
‘Help us manage tasks to serve people better’
Apacendo’s product focuses on what Lin and Sun see as the most immediate and tractable problem: daily administrative work that’s keeping clinic staff from doing anything else.
“We like to work with the system rather than completely changing it,” Lin said. “We’ve spent our entire project talking to nephrologists in every single community across the country, figuring out what they need, and how we can build technology that gets them from where they are to where they want to be.”
Their software creates “digital employees,” or AI agents that operate inside existing workflows. For example, when a fax arrives, an agent could read it, extract the relevant patient information, and upload it into a database. The company works with their partners to understand their specific pain points.
For a small practice handling around 60 faxes per day, each taking about five minutes to process manually, that adds up to around five hours of staff time recovered daily. One early customer told Lin that the tool had given her back meaningful time with her family. Porter, who uses Apacendo at Global Kidney Care, said the priority is clear.
“The biggest change we need to see is with more technology for back-office support,” she said. “We’re not looking for AI to replace people, but to help us manage tasks to serve people better.”
The startup’s ultimate goal is to use data to strengthen clinical protocols that reduce hospitalizations or delay disease progression. Identifying which patients are most likely to deteriorate benefits everyone in the system, Lin said. Amro’s hope is that eventually, technology could flag subtle signs of kidney disease and route that information to the right provider before the window for intervention closes.
“This comes back to the patients every single day,” Lin said. “We’re all going to age in this system. There just has to be a better way.”
