InFocus CXOs
“The future of AI in healthcare will not be defined by how many algorithms we deploy in the clinic, but by how intelligently we use data to understand, anticipate, and serve patient needs.”
When I am asked where artificial intelligence will most change Indian healthcare in the next five years, my answer surprises people. It is not the consultation room. It is not the radiology suite. It is the OPD funnel.
That answer reflects something I have come to believe after two decades of running hospitals across multiple groups and geographies in India. The headline AI story today is clinical. Generative models drafting discharge summaries. Algorithms reading scans. Conversational agents triaging symptoms. These will matter. But they will diffuse slowly, because Indian healthcare runs on a trust gradient that is steeper than most technology forecasts acknowledge. Patients trust their doctor. Doctors trust other doctors. Anything new must earn its place inside that hierarchy before it earns a place in the workflow. Clinical AI is doing that work. It will get there. It will simply take longer than the conference circuit suggests.
The quieter story, the one I think will compound faster and matter more to hospital economics, is happening one layer behind the consultation. It is happening in how patients find us, how they choose us, where they drop off, and what we learn from the patterns that no human in our system was ever positioned to read.
Most hospital leaders treat the OPD as an operational throughput problem. Volumes, conversion to admissions, doctor utilisation, waiting times. All real, all worth managing. But this is the wrong altitude for the AI conversation.
The OPD is the highest resolution signal of patient intent that a hospital ever receives. Every search query, every call to the contact centre, every appointment booked and not kept, every specialty chosen and abandoned, is information. Multiplied across hundreds of thousands of touchpoints a year, it is a dataset that tells you which catchments are growing, which specialties are losing relevance, which doctors are quietly building reputation, where competitors are pulling demand away, and which clinical programs the institution should invest in next.
Read at human scale, this signal is invisible. Read at machine scale, it becomes the most strategically valuable feedback loop a hospital owns.
In the networks I have led, the channels that drive the bulk of OPD growth are rarely the ones that get the bulk of leadership attention. The contact centre, digital acquisition, and community outreach quietly do the heavy lifting, while clinical excellence and brand reputation get the credit. AI does not change which channels matter. It changes how legibly those channels can be measured, attributed, and improved. That is a quieter revolution than autonomous diagnostics, but it touches every line of the P&L.
Three observations have hardened for me as I have watched these patterns across hospital groups:
The first is that demand signals travel faster than supply response. When digital acquisition softens in a catchment, contact centre conversion follows four to six weeks later, and OPD volume follows that. By the time a CFO sees the volume number, the cause is two cycles old. Hospitals that build AI to read the leading indicator, not the lagging one, are running their growth engine from a different operating tempo than peers.
The second is that the most expensive mistakes in capacity planning are made on stale assumptions about demand mix. Specialty composition shifts faster than five year master plans assume. AI applied to OPD intent data lets institutions correct course in quarters, not budget cycles.
The third, and most counterintuitive, is that the doctor allocation question is partly an AI question. Which consultant should sit in which clinic, on which day, against which referral pattern, is a complex matching problem we have historically solved with intuition and politics. Done well with data, it lifts both clinical revenue and consultant satisfaction. Done by feel, it leaves money and goodwill on the table.
None of this requires the institution to wait for clinical AI to mature. The data already exists. The question is whether leadership treats it as exhaust or as an asset.
Once a hospital accepts that the back office is the strategic frontier, the next question becomes uncomfortable. What should the institution build, what should it buy, and what should it borrow through partnership?
I would offer a simple frame. Build what touches institutional identity. Buy what is commoditised. Borrow what is strategically important but not yet core.
Build covers the patient intent platform itself. The way a hospital understands its own demand is too central to outsource. If a third party owns that intelligence, the institution becomes a tenant in its own commercial future. Build also covers anything that defines clinical experience or carries reputational risk.
Buy covers the underlying machinery. Speech recognition, large language models, scheduling optimisation engines, document intelligence. These are becoming cheap and excellent. Building them in house is rarely a defensible use of capital.
Borrow covers the spaces where speed matters more than ownership. Community first response networks, diagnostic AI partnerships, payor analytics. The right partners will be far ahead of any hospital group on technology, and the institution gains by integrating well rather than competing on infrastructure.
The most common error I see is misallocation across these three modes. Hospitals build what they should buy, buy what they should build, and borrow what they should own. The cost of that misallocation will only grow as AI capability expands.
The institutions that will lead Indian healthcare through this decade will not be the loudest about AI. They will be the ones that have made a quiet, deliberate set of choices about where intelligence sits in their operating model, who owns the data, and how the back office becomes a source of strategic advantage rather than a cost centre to be optimised.
The OPD already knows where the system is heading. The question is whether leadership is listening at the right altitude.
That, more than any single algorithm, is the AI test for our sector.
The Journey Into Industry
Biju Soman Nair is a visionary healthcare executive with over twenty-five years of experience across leading Indian hospital networks. He brings deep expertise in multi-site operations, commercial strategy, business growth, and institutional transformation.
Throughout his career, he has held cluster-level and group-level leadership roles spanning hospital P&L management, marketing, patient acquisition, and digital transformation initiatives. Known for his strategic perspective on healthcare operations and growth, Biju writes extensively on the evolving dynamics of Indian private healthcare, healthcare technology, patient engagement, and organizational excellence.
.jpg)