Kelvin Chan

A calculator nobody asked for became the whole company

Kelvin Chan spends his days on a problem most people would rather never think about. At Nirvana, the New York company he co-founded in 2019, he leads product and serves as president, pointing machine learning and data science at one of healthcare's most reliably frustrating moments: the point where a patient tries to find out whether their insurance will cover a visit, and how much it will cost.

For a company working on something this dry, Nirvana has a surprisingly personal origin. Back in 2018, Chan went looking for a therapist. The hard part was not finding one. It was learning whether his insurance would pay for it. He got on the phone with his insurer and stayed there for 48 minutes. When the call ended, he still did not have an answer.

That gap between a straightforward question and an impossible-to-get answer stuck with him. Chan and his co-founders, Akshay Venkitasubramanian and Urvish Parikh, started Nirvana to close it. Today the platform runs eligibility checks and cost estimates across digital health platforms and provider offices, with a particular focus on behavioral and mental health, where coverage rules tend to be their most tangled.

The first version of Nirvana was not the product that stuck. The team started by building back-office tools for therapists, the sort of scheduling and operations software that promised to tidy up a practice. But users kept reaching past all of it for one small, free feature: an insurance verification calculator that told them, quickly, what a patient's plan would cover. That calculator, not the scheduling or the payments, turned out to be the thing people actually wanted.

Chan has a name for the lesson he took from that. He calls it narrow excellence. “Build products that are really good at one thing and be the best at that one thing,” he says. It is an unglamorous discipline in an industry that rewards sprawling roadmaps, and it has shaped how Nirvana picks its battles: instead of trying to fix every part of a provider's operations, it works to make coverage data trustworthy enough that people stop double-checking it by phone.

From a Nobel laureate's lab to insurance data

Chan did not set out to work in insurance. He studied neuroscience at Columbia University, drawn to questions about how memory forms and how neurodegenerative disease takes it apart. He wanted into the lab of Eric Kandel, the Nobel laureate whose work on memory is foundational to the field. He got there the slow way, by showing up, visiting repeatedly, and getting to know the people who worked there until a spot opened. The persistence is a thread that runs through his story.

Academia, though, moved at a pace that did not match his appetite for action. He shifted toward industry, first into healthcare consulting at IQVIA, where he worked on financial diligence and market access for drugs, then to Enigma, a data science software company. Over four years there he rose from senior associate to director of product, building and scaling enterprise software powered by machine-learning models for fintech, pharma, and healthcare. That combination, deep data work aimed squarely at messy real-world systems, is the toolkit he brought to Nirvana.

It is a demanding standard to set for yourself. Chan argues that startups should not chase incremental gains; they exist for profound change or they do not need to exist at all. Insurance verification, for most of its history, has been the definition of “good enough,” a process built on phone trees and faxes that technically works and satisfies almost no one. His pitch is that good enough was always a choice, and that better is available if someone is willing to rebuild the plumbing.

Investors have found the argument persuasive. In March 2022 Nirvana raised more than $11 million to help behavioral health providers simplify insurance and operations. Two and a half years later, in September 2024, the company announced a $24.2 million Series A led by Northzone, bringing its total funding to $39.4 million. The round is aimed at scaling the AI-powered verification and cost-estimation work that sits at the center of the product.

A quiet operator who runs on focus

For someone building in a high-stress corner of healthcare, Chan keeps a notably calm operating system. He meditates three or four times a week, using the Insight Timer app, and starts each morning by naming the three tasks that matter most that day. He keeps a light footprint on social media, and reads widely; during one 2025 interview he mentioned working through Robert Caro's The Power Broker, the sort of doorstop biography that rewards exactly the kind of patience he showed getting into Kandel's lab.

That temperament shows up in the way Nirvana is run. Where a lot of founders equate ambition with breadth, Chan equates it with getting one thing unmistakably right and refusing to be distracted from it. The company still describes its work in the plainest possible terms: help people know what their care will cost before they receive it. It is not a flashy mission. It is a specific one, and specificity, more than anything, is what Chan seems to be optimizing for.

Ask what he is aiming at and the answer is consistent with everything else about him. He wants insurance eligibility and cost estimation to become instant, accurate, and trustworthy, so that the 48-minute call that started all of this becomes a story people struggle to believe was ever true.