As artificial intelligence technology advances, many health systems are shifting their focus from AI tools focused on individual tasks to systems responsible for end-to-end operational workflows.
Cleveland Clinic is working with startup Luminai to automate complex administrative tasks, starting with areas such as referral management. The Cleveland Clinic, one of the world’s largest academic medical centers, reported nearly 16 million patient visits in 2025. We serve patients in 23 hospitals and 300 outpatient facilities around the world, and referrals are often the starting point for their treatment.
Processing referrals relies heavily on manual review of faxes and manual interpretation of unstructured information.
“Healthcare management functions as a large manual coordination layer, and encoding that work into software has historically been difficult because workflows span systems and point solutions, rely on unstructured input, and require embedded business and clinical context at every step,” said Kesava Kirupa Dinakaran, founder and CEO of Luminai.
Recent advances in AI can now directly handle that complexity, he said, making it possible to reliably execute complete workflows rather than just automating individual tasks.
“We started with referrals because it’s one of those workflows that touches a lot of systems and requires a lot of coordination behind the scenes. There’s acceptance, validation, and follow-up. Historically, much of that process relied on caregivers manually reviewing information and reconciling it across different systems. In our pilot, we found this technology can help streamline those steps,” said Cleveland Clinic Executive Vice President Rohit Chandra, president and chief digital officer, told Fierce Healthcare.
“All of these referrals were being processed through a live fax operation. We were receiving millions of faxes. The operations organization’s job was to go through all the faxes and the handwritten notes within them, read and extract both operational and clinical data from those faxes, import it into the EMR (electronic medical record), and begin the scheduling process,” Dhinakaran said. “There are a lot of nuances there, from extracting basic details about the patient to identifying whether the patient is a high-risk patient. Does this person need urgent care? Is there a schedule available to them? There are a lot of nuances and complexities there.”
Luminai has built a virtual inbox agent that can prioritize incoming faxes and automate referral workflows. “Luminai is our first line of defense to determine whether a fax is a referral, and if so, whether it’s a high-risk, high-urgency referral. If so, we process them, extract the data, match them to the appropriate provider, and begin the scheduling process. Even if they’re not a high-risk patient, we continue to process them, move them through the queue, and make sure the issue is resolved. All that come into Luminai’s systems Faxes are processed more quickly, “which in most health systems can often take days, if not weeks,” Dhinakaran said.
He continued, “What’s interesting about that process is that at the top of the funnel, and then there’s a lot of other downstream processes like eligibility, scheduling, post-treatment appointment follow-up, revenue cycle, supply chain, etc. A lot of these very long-form process work is still happening across the administrative side where Luminai has become a platform partner for these institutions.”
The Cleveland Clinic and Luminai collaboration is moving from a pilot phase toward broader deployment.
“Advances in AI have given us the opportunity to rethink and transform many of our core capabilities. This is one of many capabilities we can do over time. It was about supporting real operational challenges. As we saw the results in a pilot environment, we became more confident that this could be scaled in a useful and sustainable way.
While many health tech and AI companies focus on narrow use cases such as scribing and revenue cycle management, Dhinakaran said Luminai aims to become a management operating platform partner for large healthcare systems.
The company was founded in 2020 and signed its first health system customer in 2023. Luminai currently works with 20 large healthcare systems. Earlier this month, the AI-native automation platform closed a $38 million Series B funding round led by Peak XV Partners (formerly Sequoia India &Southeast Asia). The round includes participation from new investor Define Ventures and continued support from existing investors including General Catalyst and Y Combinator.
Luminai’s platform combines healthcare-trained AI models, a configurable workflow engine, and human validation. The company is using the new funding to expand product capabilities, expand its engineering and implementation teams, and support additional enterprise customers, executives said.
Dhinakaran grew up in India and was a professional Rubik’s cube solver in his youth, holding the Guinness World Record for most Rubik’s cubes solved in one hour.
“When I first encountered process automation problems, it felt very familiar to me in that these were 80-step problems that needed to be done in five steps,” he said.
About eight years ago, Dhinakaran moved to Silicon Valley and began exploring and experimenting with AI to automate complex tasks. When he was exposed to the American health care system, he was surprised to find that operations are still primarily based on people, processes, and paper. “As I walked through the administration buildings of some of these hospitals, I was just struck by the amount of repetitive, very manual, very operational work that people were doing,” he said.
“At the end of the day, there are a set of 8 to 10 core work streams that perform health system management. These core work streams are actually quite interconnected, but today there are hundreds, if not nearly 1,000, point solutions that solve these individual problems. The average U.S. health system has more than 400 points for these 8 to 10 core work streams. I think there is a solution,” he said. “All of these point solutions are very difficult to maintain because it takes time to set up the infrastructure, give it context, and actually start automating and operating it.”
Dhinakaran argues that the technology industry has reached a “cycle of intellectual abundance” where it is finally possible to build deep and broad AI platforms. “What we’ve learned is that the main reason we haven’t been able to do that in the past is because there’s a huge amount of unstructured data out there. So unless you structure this data effectively and in a scalable way, you can’t actually build a software layer to do this work,” he said.
Luminai brings together applied AI talent with experience from organizations such as Palantir, Cruise, Google, Coinbase, and Brex, as well as healthcare professionals and product leaders from Epic, Banner Health, and other large care delivery and health IT environments.
“It takes a long time to build healthcare. To be successful, you have to be very focused on product and engineering and have deep customer penetration. We’ve basically spent the last two-and-a-half years doing that,” Dhinakaran said.
Rather than automating one step at a time, health systems are increasingly looking to technology platform partners that can deliver AI to run multi-step workflows across systems and support a wide range of long-term high-impact use cases, Dhinakaran argues.
Administrative tasks still account for up to 25% of healthcare spending. Large health systems desire to automate administrative tasks as they face increasing cost pressures, staffing constraints, and increased operational complexity.
As Cleveland Clinic works with Luminai, Chandra said, technology leaders see an opportunity to use the company’s technology to address other administrative workflows across the health system, particularly high-volume, complex, and operationally intensive workflows.
“As we ponder where to go next, our broader focus is on areas where teams are spending too much time on repetitive administrative tasks and where increased speed and consistency could have a meaningful impact,” he said.
“While we are starting to see a shift towards institutions partnering at the platform level, it is critical that it is rooted in the reality that we can actually ensure ROI for every use case we choose to deploy,” Dhinakaran said. “That’s what we’re all about: Let’s solve real problems. Let’s actually drive automation and task elimination in areas where real money is being spent today and move that to software and AI.”
He added, “The magical experience of tinkering with AI systems is starting to wear off, and we’re at a point where we need to do the work of applying these to real-world use cases with the nuances of how we operate them. There’s a real need in these health systems, and we’re interested in facilitating that kind of work that actually impacts the bottom line.”
Chandra argues that automating complex workflows such as referrals not only improves operational efficiency, but also benefits healthcare providers and patients.
“Administrative workflows have a greater impact on the patient experience than many people realize. When those processes run more efficiently and reliably, patients realize it. “It also supports a more responsive experience by ensuring the right information is captured and each patient’s needs are appropriately addressed as they move through the system.” “Providers can feel the difference, too. A smoother administrative process means fewer interactions, less operational burden, and less time for teams to spend making manual adjustments.”
According to Shailendra Singh, managing partner at investor Peak XV Partners, Luminai is building an intelligent orchestration layer that will define the capabilities of future healthcare operations.
“The company’s engineering rigor and customer-built execution model make them the foundational infrastructure as health systems fundamentally rethink how they perform operational tasks,” Singh said in a statement.
