How Sevagram Went Digital

In late 2003, when the idea of a hospital information system first landed in Sevagram, it did not arrive with fanfare. It came in the form of official letters, unfamiliar names, and a promise that sounded almost too ambitious for our setting. At that time, MGIMS was a busy rural teaching hospital running largely on paper, memory, and human stamina. Our wards were full, the OPDs overflowed, and every department had its own way of doing things. Files moved from counter to counter like tired messengers. Discharge cards were handwritten in a hurry, lab reports travelled in envelopes, and billing depended on slips that could vanish in a pocket. We were not dreaming of “digital transformation.” We were simply trying to survive the daily load without losing information—and without losing our minds.

The Center for Development of Advanced Computing (C-DAC), an autonomous scientific society under the Government of India, had experience building health software. They had implemented a hospital information system at SGPGI Lucknow, and they had attempted similar work at two Delhi hospitals with mixed results. Still, the government decided to take a chance on them again, and Sevagram became the testing ground. The project was formally titled “Design and Development of Advanced Hospital Management System and Prototype Implementation for MGIMS at Sevagram, Maharashtra.” The phrase “prototype implementation” mattered. It meant this wasn’t the final grand system; it was a working model that would be built, tested, corrected, and only then scaled up. In plain language, we were going to be the place where the system would learn to walk.

On 30 December 2003, Mr R.K. Verma, Executive Director of C-DAC Noida, wrote to the Department of Information Technology and identified the team: Mr Naveen Kumar Jain as Project Director, Mr Pradip Parida as Project Manager, and Mr Praveen Srivastava as Senior Project Engineer. A few weeks later, on 12 February 2004, Mr P. Soreng, Deputy Director in the Ministry (MeitY), formally tasked C-DAC to design and implement the HIS at MGIMS. The budget was generous for that time—₹100 lakhs for Phase 1, with additional allocations planned for Phase 2 and Phase 3. The timeline, on paper, was tight: March 2004 to February 2005. In reality, like most hospital projects, it ran on human time, not file time. Extensions happened. Deadlines shifted. Eventually, MGIMS took ownership of the system in June 2008.

The plan on paper, and the hospital in real life

C-DAC proposed to build the HIS in three phases. Phase 1 covered the clinical spine—registration, appointments, OPD, IPD, operation theatre, labs, blood bank, diet, CSSD, pharmacy, EMR, and billing. Phase 2 expanded into the less glamorous but equally essential hospital machinery—sanitation, laundry, security, transport, accounts, procurement, stores, and inventory. Phase 3 aimed at what sounded futuristic then: PACS for radiology and other imaging-heavy departments.

On a PowerPoint slide, this looked neat. In a hospital corridor, it looked like chaos waiting to happen. Every module had to speak to another module. Every department had its own habits, its own shortcuts, and its own resistance to change. Most importantly, our staff had not grown up with computers. Many were nervous even touching a mouse. We had no culture of “data entry.” We had a culture of “somehow manage.”

And then, even before the project truly began, we suffered an early setback that would have discouraged a less stubborn team: the initial project manager left before completing the requirements gathering. It felt like losing a wicket without scoring a run. The first ball had not even been defended, and we were already walking back to the pavilion.

The consequence was painful. Hospital staff had spent time explaining workflows, showing registers, walking engineers through OPDs and wards. That information—the first and most precious “requirement document”—was lost. When the new team arrived, they had to begin again. But the second time, the energy was lower. People were polite, but less enthusiastic. Some were openly skeptical. A few had already decided this would fail, and they were only waiting for proof.

This is where Mr Praveen Srivastava became crucial. He stepped in as project manager and held the work together with patience and grit. He was in his late thirties then, trained in mathematics and computer science, and had joined C-DAC as a senior engineer. He had the calm of a man who knew projects don’t move forward in straight lines. He regrouped his team, started collecting requirements afresh, and slowly brought structure to what could have easily become a government-funded mess.

Years later, he would look back on Sevagram as a turning point in his own career—his entry into the health domain, his growth as an expert, and the eventual success of e-Sushrut as a widely known HMIS product. But in 2004, none of us were thinking of future reputations. We were thinking of the next day’s OPD.

The engineers arrive in a village hospital

A hospital information system is built by software, yes—but it is implemented by people who show up every day and take abuse with a smile. Sevagram saw a steady stream of C-DAC engineers who lived among us, ate our food, sweated in our summers, and learned our hospital’s language the hard way.

The early team included Rajeev Yadav, Ajay Gupta, Devendra Rao, Pradeep Kumar, Jitendra Dubey, Ravi Sinha, and later several others. Some stayed on campus for long stretches. Three resident engineers—Ashish Singh, Mantosh Kumar, and Amit Kumar Digaliya—spent extended periods in Sevagram, almost like rotating house officers, except their patients were servers, printers, and angry users.

Each engineer developed a personality in our collective memory. Ashutosh oversaw cabling and networking. Ravikumar built the insurance module in a small pharmacy room. Ajay Gupta handled the medical and central stores. Rajeev Yadav worked on student, personnel, and accounts modules. Mantosh arrived later and did what good troubleshooters do—he fixed what others had left half-broken. Ashish Singh, the resident engineer from September 2004 to July 2006, became our bridge to Noida, keeping the umbilical cord alive.

Amit Kumar, deputed in late 2006, became something more than a site engineer. He became part of the extended Sevagram family. Hemant, one of our own young staff members, remembers taking him around the villages, introducing him to local food, and making him feel at home. Amit would visit homes, taste Maharashtrian meals with the curiosity of a student, and sometimes drive the team to nearby places like Bordharan and Mahakali. Those small outings mattered. They turned a “project” into a shared life.

One episode from 2005 still makes me smile. Two young female engineers came to Sevagram for a short stay. They were warned—lightly, casually—not to step out during Holi. They did not take it seriously. Sevagram Square, on Holi, does not do “light colours.” It does full immersion. They were drenched, startled, and suddenly anxious in a crowd that meant no harm but looked intimidating. They called Hemant, who rushed out, escorted them safely to their guest house, and arranged food and water because the town had shut down for the festival. It was a small act of care, but it revealed something important: technology may be modern, but implementation still depends on old-fashioned human decency.

My own exit, and Rajnish stepping in

In August 2004, I left for Berkeley for my ten-month MPH program. It was a good academic decision, but it created a practical gap. HIS needed daily supervision, and I could not do it from California.

I handed over the responsibility to Dr Rajnish Joshi. He was only thirty then, a young lecturer in Medicine, but he had the temperament for complex work and the ability to talk to people across departments. He had already worked with me on research projects, and I trusted his judgment. His brief was simple, almost brutal in its clarity: Get HIS on hospital desktops by August 2005.

One year. A hospital that ran on paper. Hundreds of staff members who had never used computers. A software team still learning the terrain. And the daily burden of patients that did not pause for “implementation.”

Rajnish did what good leaders do. He built a team.

Bhavana and the courage to begin again

The first and most important member of that team was Bhavana. By then, she had already written FoxPro programs and created electronic discharge summaries for our Medicine department. Few people noticed those early efforts, because hospitals are strange places: they only respect a system once it becomes unavoidable.

Bhavana’s role was not just technical. It was emotional labour. She had to work in an environment where computers still made people uncomfortable. She had to handle family responsibilities, children, and the ordinary chaos of life—while taking on a project that demanded patience every day. The fact that she did it without drama is, to me, the most “Sevagram” part of this story.

Yogesh Khond: the man who could fix anything

Every large project has a quiet hero who doesn’t appear in official letters. For us, that was Yogesh Khond.

Yogesh’s life had taught him endurance early. After his father died of oral cancer, he worked as an attendant in the Dean’s office and endured humiliations that would have broken many young men. He left his BSc halfway because survival came first. But he had a stubborn love for computers. He trained in Pune, opened a DTP shop near Sevagram Square, typed nearly a hundred MD theses, tutored students in English and Maths, and took up short projects at MGIMS wherever work appeared.

When Bhavana brought him into HIS as a network engineer in July 2004, he found his true stage. Yogesh became the person everyone called when something didn’t work—which, in the early days, was almost every hour. Hardware, networking, printers, cables, switches, software engineers, angry clerks—Yogesh handled all of it with the same steady expression. He was not formally trained as a hospital IT professional. He became one by necessity, learning faster than the system could break.

A team built from ordinary people

As the project grew, the HIS team expanded: Mahendra Chaudhari, Nitin Shende, Amit Bijwar, Sudhir Tamgadge, Vinay More, Ram Khedikar, Prashant Khond, Saurabh Nagpure, Hemant Ghumde, Rohit Meshram, and others.

Many were young. Some had arts degrees. Some had only short computer training. Almost none had experience in data collection, coding, or healthcare workflows. Yet they learned by walking—literally—through wards and OPDs, watching patient journeys, understanding how blood samples moved, how bills were made, how medicines were issued, how reports were filed.

Nitin Shende, only 26 when he joined, began by entering personal data of over a thousand staff members. Later, he visited wards to solve dot matrix printer problems and barcode issues—small tasks that kept the system alive. Amit Bijwar, recruited for multitasking, came with basic training and the humility to learn. Hemant, barely 20, transferred from the MS office, brought something more valuable than experience: common sense and teamwork.

These were not glamorous roles. Nobody applauds a young man for fixing a printer in a ward. But HIS survived because such small repairs happened every day.

Cables, servers, and the “secret” work

The physical backbone of the system was networking. In 2004, this itself felt like surgery. Cables had to connect the hospital, the Dean’s office, and the old hospital blocks. Hi-Tech Computers Nagpur was assigned the job. Fibre optic cables, CAT-5, PVC pipes, ports, racks, switches—these were new words for many of us.

One detail still amuses me: digging up the road to lay cables had to be done quietly, overnight, almost like a village heist. The contractor warned that formal permissions would not come easily, so the work happened in the hours when the campus slept. By morning, the road looked innocent again, while the cables beneath it carried the future.

Computers were installed across the hospital—OPDs, wards, labs, accounts, pharmacy, casualty. In October 2004 alone, 130 computers went in, including some Linux-based machines. For a place that had lived with registers and carbon copies, this felt like a small invasion.

A training centre was created using old systems. From May to November 2004, nearly 300 staff members received basic computer training. Many came reluctantly. Some came with fear. A few came with quiet pride, like schoolchildren learning a new language.

The insurance module: a rough beginning

The first major live module was the health insurance system, launched around November–December 2004—crucial months because MGIMS issued insurance cards only during that window.

The module went live under pressure, but the system was not ready. The application server was incomplete, so the insurance module ran on a separate server using existing PCs. Errors followed: duplicate cards, wrong cards, cash collection mismatches. At one point, there was a discrepancy of nearly ₹2 lakhs. Some of it was program malfunction, some of it was inadequate monitoring, and all of it was a warning: computerisation does not automatically create discipline. It only exposes the lack of it.

The most alarming discovery was that the source code of the insurance module could not be found. Later, during implementation work in early 2005, one PC that contained the source code was accidentally overwritten. The code was lost—again. It sounds unbelievable today, but in those early days, we were learning not just software, but software culture: backups, version control, accountability.

C-DAC promised to rebuild the module before the next cycle. We continued, bruised but not defeated.

The slow birth of a system

By early 2005, teams were working on pharmacy, investigations, billing, OPD, and ADT modules. Servers arrived. Oracle database servers came in. Red Hat Linux was installed. Red Hat Cluster was introduced with an explanation that made sense even to a non-engineer: like friends pulling together—if one drops out, the others keep the rope moving.

Even then, the servers struggled. A busy teaching hospital does not behave like a small demo unit. Transactions crawled. Staff complained. On 11 August 2006, I wrote to C-DAC with a line that came from the heart: “We look forward to a server that would run, not crawl.” It was half complaint, half prayer.

Over time, upgrades happened. RAM increased. Processors were added. Hardware improved. In 2012, MGIMS moved from Pramati to Tomcat to make the system more stable and browser-independent. Engineers from outside firms came, stayed in Sevagram, worked late into the night, and made it happen. The HIS kept evolving—not because it was perfect, but because we refused to let it die.

What I remember most

When people ask me how HIS began in Sevagram, they often expect a dramatic answer—a visionary moment, a bold speech, a single hero.

But that is not how it happened.

It began with ordinary people learning unfamiliar things. It began with engineers walking through crowded OPDs, trying to understand a rural hospital’s rhythm. It began with staff members overcoming fear of a mouse. It began with cables laid under roads at night. It began with mistakes—lost documents, overwritten code, mismatched cash, slow servers. It began with frustration, and then stubborn persistence.

And slowly, almost quietly, it became real.

The first time I saw a patient’s details appear on a screen without anyone searching for a file, I felt something shift. Not pride. Relief. The kind you feel when a long-standing headache finally eases.

Sevagram did not become digital because we were advanced. We became digital because paper had started failing us, and we were willing to learn—one module, one mistake, one night at a time.

---