It is not a technology problem. It is a design problem.

Most hospital software in India was built for administration. Billing, registration, pharmacy — those parts work. What the software does not understand is how a doctor thinks, or how a patient actually moves through a hospital.

Take a patient with diabetes. They come in, get medicines, and leave. A month later, the same patient is back. There is no system flagging that they missed their last visit, that their blood sugar has been climbing, that they are heading somewhere bad. That information exists somewhere — in a paper file, in someone’s memory, in a lab report nobody linked to anything. When the staff member who remembers that patient is off duty or overwhelmed, the information might as well not exist.

From a public health point of view, that matters. Chronic diseases need more than a prescription. They need someone tracking what is happening over months, noticing when a patient drifts, acting before the complication. Our systems were built for encounters, not for trajectories.

And the data to do better is already there. Visit history. Lab reports. Prescriptions. Patient demographics. In most hospitals, even smaller ones, this information sits in the system doing nothing useful.

A basic tool could change that. Flag patients who missed follow-up. Send a WhatsApp message. Show a doctor the trend from the last six visits in ten seconds instead of thirty. None of this requires anything exotic. It requires someone to actually design for the clinical workflow rather than the billing workflow.

Workflow is its own problem. A patient moves from registration to the doctor to the lab to the pharmacy. The software does not move with them. Each department operates separately. Staff patch things together manually, patients stand in wrong queues, time gets lost at every handoff. Better queue management, clearer patient directions, basic integration between departments — these are not ambitious ideas. They are things that should already exist.

What I keep coming back to, after working across both a large tertiary hospital and a smaller setup, is that the technology is usually not what fails. What fails is that nobody designing the software spent time watching what actually happens. The system was built for an orderly hospital. Actual hospitals are not orderly.

Any tool — AI or otherwise — will be abandoned if it adds steps during a busy OPD. A dashboard that takes four clicks will not get used. A reminder system that needs manual configuration every week will be turned off within a month.

Before building anything, the question is: what problem does this doctor or nurse actually have, and does this solve it?

Right now, hospitals keep getting more software and fewer solutions. That gap will not close until someone asks that question first, and takes the answer seriously.