For Puneet Soni, building a product is second nature. The former product manager from Google, and ex- CPO at Flipkart has overseen and built several successful ideas to scale. After his Flipkart stint, Punit was certain about one thing — he would find a way back to India either as a market for a solution he built or he would develop a service that suited the market.
Today he is the founder of San Francisco, Bay Area-based healthcare startup Suki.AI, a voice-based digital assistant for clinicians.
Founded in 2017, Suki.AI, Punit aims to build a one-stop platform for doctors, like what Google is for search engines.
Why healthcare?
“How come healthcare that contributes five to seven percent of every country’s GDP is ever needed, and larger than social networking, e-commerce, or search has a single tech company close to size of Google, Amazon or Facebook?,” asks Punit. He believes the largest tech company in the world has to be in healthcare.
Punit says he was motivated by a few different things. “My stint in India taught me the value of entrepreneurship. One is I wanted to learn and grow from a different perspective and the journey is very different from when you build things from scratch, with nothing except an idea, and build it from there. I also wanted to leverage some of the advantages I had, and more importantly I wanted to build something that had clear value to the society,” he adds.
The building blocks
Connecting the dots in the technology ecosystem led Punit to look at healthcare. Since he had no experience in the sector, he decided to dig deeper to understand the sector’s challenges. During his research, he spoke to people at the Harvard Medical School and other medical schools in the East Coast region.
“I would just go there and shadow doctors from morning until evening. I would spend the day sitting in hospitals and looking at how people were looking at things. There would be interesting stories. But what became clear throughout was that technology isn’t serving the doctors,” adds Punit.
During his hospital visits, he saw physicians dealing with burnouts, which impacted patient care. “The doctors are overworked and there are few doctors for a large number of patients. There is no technology, and for every one hour of clinical work they are doing two hours of administrative work,” says Punit.
Starting with Notes
According to Punit, while the Indian healthcare system has unique pain points, the core issues remain the same.
Punit believes that voice, which finds itself at the same juncture as mobile technology was in 2004, is beginning to have some good use cases.
“But I contend that the next big UI revolution is going to happen with voice. Typing and swiping won’t be a significant use case in the near future as voice will be,” adds Punit. It led to the birth of the idea of building a voice-based digital assistant for doctors.
“Think of it like a Siri, Alexa, or Google Assistant for a doctor. Initially AI isn’t smart. It can’t help a doctor by doing diagnostics or looking at a patient history etc. What it can do is help the doctor create notes. Now these notes are clinical documentation made by the doctors for every patient they see,” explains Punit.
These clinical notes are taken in a particular way — SOAP (subjective, objective, assessment, and plan). Subjective is what the doctor hears from the patient, Objective is what they observed, Assessment is what did they assess and plan is next steps.
“This has been the process for decades now. The first thing the Suki AI does is write a clinical note, and the doctor uses a command to get the notes. The doctor tells Suki to create a note for a patient. The app creates the note, goes into the electronic medical record of the patient, repopulate the note and keep it ready,” explains Punit.
The workings
The app helps doctors insert patient notes. The data is then put in the overall EMR (electronic medical record). This process helps save 50 to 70 percent of the time taken on documentation, helping doctors focus on other tasks.
The team has built the internal voice recognition and command system, which is the basis of Suki Speed Service, a platform to add voice to their product. The product is used by 91 different health systems in the US.
The idea behind the product, explains the founder, was to get the assistant to create notes. Most data in healthcare is created by doctors, so the act of creating these notes is a step in training the AI to be smarter. Initially, the app worked as a basic note-taking platform. It is now evolving and provides other services such as writing diagnostic information, problem-based charting and adding coding information, helping with patient engagement etc.
“This way you slowly start rethinking the entire tech stack and making it invisible and assisted. This could be a long journey,” adds Punit. The idea is to help doctors to focus on their clinical work. Punit explains even with tech growing and exploding in regions like India, there still is a long way to go.
Market and future
The conversational AI market is aimed to touch $18.6 billion by 2026 according to research firm Markets and Markets. Some of the Voice based companies include – Slang Labs, Alan AI and Houndify that offer conversational voice platforms. Healthcare startups in India includes Practo, Mfine, among others.
The healthcare industry world over is highly regulated. For every one doctor, there are three administrators in the system. But changing this, Punit says, can happen only through technology.
Suki.AI’s revenue model is simple. Doctors can use the Windows, IoS, or Android app and subscribe to a pay-per-month plan. It follow a SaaS (software as a service) model. While the actual pricing has been undisclosed, the team stated they are now working with one of the largest tele healthcare companies in the US.
“We are now working to transform the tech in the healthcare industry,” says Punit.
