From fax machines to AI: Medi Assist bets big on transforming Indian healthcare sector

Not too long ago, if you tried to file a health insurance claim in India, there was a decent chance your hospital was sending discharge bills through a fax machine at best or physical copies at worst. Not exactly the image you’d expect from a country marching steadily into a digital future. Enter Medi Assist, an unlikely hero in an industry plagued by archaic processes. By combining AI with digitisation, and a singular focus on user experience, this company has quietly started to reshape India’s healthcare insurance sector unbeknownst to most of us.

“It’s been a long journey,” recalls Medi Assist CEO Satish Gidugu during our interview conversation, clearly amused by how things used to run when he joined the company over a decade ago. “When I joined this organization 11 years ago, hospitals used to fax patients’ discharge bills to us, right? And I have actually seen fax machines pushing out these invoices!” The wonder in his voice says it all.

Also read: AMD’s Jaya Jagadish: India to play ‘transformative role’ in global AI, datacenter and chip landscape

Once burdened by manual processes and outdated technologies, India’s healthcare insurance ecosystem is now poised for a radical transformation. And Medi Assist, a once-ordinary Third Party Administrator (TPA), is one of the few companies aspiring to disrupt the status quo.

Indian healthcare sector is ready for reinvention

India’s health insurance ecosystem is notoriously fragmented. Think of it as a giant labyrinth with multiple insurers, thousands of hospitals, each with different forms, disclaimers, and haphazard record-keeping. “The biggest challenge in Indian healthcare is that there are no common standard data formats anywhere in the country,” Satish laments. “The hospital policies are written in Word documents… lots of conditions, lots of policy language that needs interpretation.”

In simpler terms, it’s a data nightmare that can overwhelm the best of us. Hospital records? Could be an Excel sheet, a scribbled note, or a typed-out PDF. Policy documents? Possibly a 10-page Word document with complicated clauses and sub-clauses only a lawyer could understand. If you’ve ever filed a claim, you know how that chaos can chew up time, money, and sanity – all of which are in short supply during a medical emergency.

Medi Assist wants to tackle this bottleneck head-on. Over the last four years, they poured resources into proprietary digitisation methods, scanning and standardising millions of bills. “We use a fair bit of machine learning models today to detect fraud and, you know, prevent those shady claims from getting paid,” Satish explains. The Medi Assist tech stack, however, is about more than just scanning documents –  it’s about creating a rules-based algorithmic engine that can interpret unstructured policy text, check it against the scanned hospital data, and instantly figure out which charges are valid and which are suspicious.

AI for fraud detection in hospital claims

The real magic, as you might guess, comes from artificial intelligence. Dhruv Rastogi, the Senior Vice President & Head of Data Science at Medi Assist, came aboard with experience spanning Vodafone, Reliance, and Nomura. “I have been working in data science, AI data-plot platforms for quite a while now,” he says. “I initiated scaled data science teams and drove a lot of growth and digital transformations for organizations leveraging AI and data.”

At Medi Assist, Dhruv’s AI capabilities power some of the most cutting-edge solutions in the insurance TPA landscape, of which fraud detection is a big chunk. Historically, detecting a suspicious claim was a tedious manual process – someone might glance at a weirdly high invoice or notice that the same hospital bill came in for two different patients. But those manual processes catch only the obvious red flags, not the more subtle and ingenious fraud manifestations.

Also read: ARAS with AI for two-wheelers: How Luna Systems aims to save lives on Indian roads

And the result speaks for itself, according to Satish. “The savings that we’re delivering on account of frauds to insurers from February 2024 till now, over the last 6 months, have doubled. This is simply because we’re able to bring over 160 parameters from all the data sources, find models and correlations that humans just can’t find,” Satish emphasises, underscoring just how much the proprietary AI engine developed by Medi Assist is accelerating the delivery of value to the healthcare ecosystem.

Building a tech platform that enhances transparency

Amid all this talk of invoice scanning and AI, Satish highlights another critical piece of the healthcare journey – which is building an interface that benefits not just the insurer or hospital but also the end user, the patient. 

Medi Assist insists that every transaction is mirrored in real time across everyone’s dashboards. “The hospital, the insurer, the agent, the member, everybody and us, we see the same thing in real time,” according to Satish Gidugu. In an industry infamous for secrecy and finger-pointing, needless to say this transparency is a breath of fresh air.

Of course, to make all of this work, there’s also scale involved in the equation. “We work with more than 12,000 corporates… more than 30,000 hospitals, all the major health insurance companies as such,” Dhruv Rastogi notes. Integration at that level requires robust data governance and standardized access controls, he notes. But once that’s in place, you unlock the potential for real-time analytics, faster reimbursements, and an overall sense that maybe health insurance doesn’t have to be painful.

Intelligently cutting down hospital discharge time

One direct perk of all the advanced AI baked into Medi Assist’s tech stack is a predictive model that gauges a patient’s out-of-pocket expenses before they’re even given the final bill. “We spent the last couple of years working on this feature, and had multiple iterations and versions to enhance its accuracy,” Satish says. The result? Some 7,000 patients have been discharged quickly, without the usual hours-long suspense waiting for final payment clearance. That’s a big deal for patients who might otherwise spend unnecessary hours in hospital premises, worrying about medical costs or worse.

Also read: Osmo’s scent teleportation: How AI is digitising scent for the first time

“In the end, it’s about building that frictionless moment of truth, so you can just walk out,” Dhruv adds. The AI basically cross-references your claim, hospital tariff, policy coverage, plus any past data to project the final tab. That shortens wait times, cuts overhead for hospitals, and fosters a sense of trust that’s historically lacking in the Indian medical insurance and healthcare sector at large.

Additionally, with so many tie-ups with medicare providers, Medi Assist processes nearly a quarter of all health insurance claims in India, Satish says, and “85 percent of hospital-submitted claims processed through an electronic workflow,” he emphasises. If you think about the ripple effect, that’s thousands of hours saved daily across hospitals, insurers, and patients and their families.

But according to Satish, it’s not just about digitising a failing system – it’s re-engineering the entire claims workflow. By standardising data, training AI to handle everything from suspicious billing codes to minor coverage details, and enforcing real-time transparency, they’re arguably turning an archaic ecosystem into a modern, data-driven enterprise of the 21st century.

Building the future of Indian healthcare

Both Satish and Dhruv sound bullish about the future, when I asked them about looking ahead at what the future of Indian healthcare might look like. With the Indian government’s push for digital identity (ABHA IDs) and open data initiatives, the stage is set for deeper integration, they say. Think of an environment where your coverage data merges with your telemedicine app, your e-pharmacy, and possibly your wearable device. That’s when insurance stops being a hush-hush black box of complexities and denial, and evolves into a truly personalised, dynamic system.

“Can our health insurance plans be customised and personalized?” wonders Dhruv, pointing out that technology can easily differentiate your risk profile from someone else’s. If you’re someone who never touches a cigarette, hits the gym regularly, or has a family history that suggests minimal risk, why not a plan that reflects your lifestyle? That might finally turn health coverage from a reluctant purchase into something that actively invests in your well-being.

And to kickstart that process, Medi Assist along with Boston Consulting Group (BCG) unveiled a groundbreaking framework for “Borderless Health” in November 2024. Still at a conceptual stage, the Borderless Health framework is Medi Assist’s vision to eliminate healthcare disparities by creating a data-driven, technology-enabled model for delivering healthcare services. At the core of this new approach is the “Health JAM Trinity,” which consists of three key components – Joined Health Data, Automation, and Mobile-Enabled Healthcare. 

These pillars will enable more efficient, equitable, and accessible healthcare services across India, according to Medi Assist. “Achieving universal cashless healthcare by 2047 will require coordinated efforts and collaboration among healthcare providers, insurers, employers, and policymakers,” emphasises Satish.

As I reach the end of my conversation with Satish and Dhruv, I can’t help but wonder at Medi Assist’s story being a testament to how even the most deeply entrenched, paper-heavy industries can transform with the right blend of AI, data, and willpower. 

“I don’t know how you do your job, protecting the milk and keeping the cat happy,” jokes Satish, recalling a comment made by a longtime industry observer. But that’s precisely the challenge Medi Assist has taken up, of bridging the needs of insurers, hospitals, and patients with minimal friction. And the difference it could make for millions of Indians who rely on insurance not just to cover a bill but to finally offer peace of mind.

Also read: Meet Prithvi, NASA & IBM’s free AI model for better weather prediction