One in three patients prescribed a specialty drug never starts treatment. Not because the therapy does not exist, but because the systems designed to support their access are too slow, too fragmented, and too disconnected from the people they are meant to serve. It is a problem that sits at the intersection of enterprise technology and human health, and it is one that data architect Sujit Murumkar has spent the better part of two decades trying to solve.
Murumkar is a technology and data engineering leader whose career spans pharmaceutical giants, Fortune 500 companies, and some of the world’s largest financial institutions. His focus has never been technology for its own sake. It has been what technology enables, specifically whether a patient with a newly diagnosed chronic condition gets their medication within days or weeks, whether a physician receives accurate information before making a treatment decision, and whether a healthcare organization can act on emerging signals before they become clinical crises.
When Data Gaps Become Patient Gaps
The commercial infrastructure behind a pharmaceutical product is more complex than most people outside the industry realize. Before a therapy reaches a patient, it passes through layers of systems handling physician targeting, payer access, patient enrollment, adherence monitoring, and field team coordination. When those systems operate on outdated data, siloed platforms, or batch reporting cycles measured in days, the delays are not just operational. They are clinical.
“A field representative who learns about an access barrier three weeks after it occurred cannot help that patient,” Murumkar says. “The data has to be current, connected, and trusted before it becomes useful to the people making decisions that affect patient care.”
Research published in the Journal of Medical Internet Research confirms that real-time digital health data systems directly improve clinical outcome measurement, enabling earlier intervention and more responsive care coordination. The gap between organizations that operate on real-time data and those relying on legacy batch systems translates into measurable differences in patient engagement rates, adherence outcomes, and ultimately therapy effectiveness.
Murumkar led the development of commercial data platforms at a Fortune 100 pharmaceutical company that unified patient services, field force operations, market access, and marketing analytics into a single, governed data environment. The result was a reduction in reporting cycles from several days to a matter of hours, giving commercial teams the kind of immediacy previously available only in financial trading environments.
Connecting Dots Across Disconnected Systems
One of the most persistent challenges in pharmaceutical patient support is that the data relevant to a patient’s therapy journey sits in entirely separate systems. Prescription data lives in one place. Insurance adjudication records live in another. Field team notes, patient assistance program enrollment, specialty pharmacy fulfillment, and adherence tracking each occupy their own siloed infrastructure. Without a technical architecture that connects these domains through shared patient identifiers and unified data models, organizations cannot see the complete picture.
Murumkar addressed this directly by designing global master data management models for customer and patient domains. These models standardized how patient records were identified, matched, and maintained across business units and geographies, enabling analytics teams to build a coherent view of the therapy journey from prescription to patient adherence.
“When you can trace a patient’s journey across every touchpoint, from the moment a physician writes a prescription to whether that patient refills in month three, you can actually identify where the system is failing them,” he explains. “That requires more than good intentions. It requires architecture that was designed with that outcome in mind.”
The platforms his teams delivered enabled next best action models for patient support teams, real-time omnichannel performance visibility, and AI-driven identification of patients at risk of abandoning therapy. Across several major pharmaceutical clients, these capabilities contributed to measurable improvements in patient engagement and a 60 percent reduction in reporting lead times for commercial operations.
AI at the Point of Care
Predictive analytics is reshaping how pharmaceutical companies approach patient support. Rather than reacting to adherence failures after they occur, organizations with mature data infrastructure can identify risk factors weeks in advance. A patient who fills a prescription but delays a refill, encounters a prior authorization rejection, or stops engaging with a patient support program sends signals that, when captured and analyzed in real time, allow for proactive outreach.
Murumkar’s work in building AI-enabled insights orchestration models created the technical foundation for these capabilities. By standardizing feature engineering across commercial, medical, and patient services domains, his teams reduced the time required to deploy new predictive models by up to 60 percent. This acceleration matters in environments where drug launches move quickly, and patient support programs must scale from hundreds to tens of thousands of patients within months.
“Building one model for one use case is relatively straightforward,” Murumkar notes. “Building a data platform where models can be deployed, monitored, retrained, and scaled across multiple therapeutic areas and markets is an entirely different engineering challenge. That is the work that actually changes outcomes at scale.”
From Infrastructure to Impact
What separates Murumkar’s approach from conventional IT modernization is the explicit connection he draws between technical decisions and patient results. Data governance frameworks, master data models, and cloud migration strategies are rarely discussed in terms of human health outcomes. In the pharmaceutical context, however, every architectural decision carries downstream consequences for the people the industry exists to serve.
His career spans work at organizations managing multi-trillion dollar portfolios, global life sciences companies operating across dozens of regulated markets, and patient services organizations responsible for helping people navigate complex access barriers. Across all of these environments, the core challenge has remained consistent. Organizations collect more data than they can act on, and the infrastructure needed to close that gap between collection and clinical utility requires both technical depth and a clear understanding of the human stakes involved.
For an industry where the distance between a data pipeline and a patient outcome can be measured in lives, the case for intelligent data architecture is not primarily financial. It is fundamentally about whether the systems organizations build are worthy of the mission they claim to serve.
