In her view, agentic AI and the like are what will move the needle for outcomes and speed in Indian healthcare. It’s a push that makes sense given how fast enterprises are latching onto AI and the kind of data security they’re being asked to prove these days.
Bhattacharya, who runs South Asia for Salesforce, puts it in plain terms: there are still no cures for a lot of things and the path to care is often a hassle. But with AI, robotics and some automation, you can make a patient’s journey a little less of an ordeal and see some real returns.
Healthcare’s AI moment
It comes down to two voids. For one, we only have medicines for a third of diseases, so pharma has to keep up. For another, too many processes are needlessly invasive or just don’t work well. AI can help you find your way around both of those.
Take surgery: where a 15-day stay was once the norm, better tech has you out in three. And while that’s happening, agentic systems are there to handle the legwork of appointments and referrals so staff can focus on the care itself.
India strategy moves to localise AI
Salesforce is putting together an India-specific plan to make this happen. You’ll see MuleSoft on Hyperforce coming to the region, for instance, so companies can run their integrations and AI workflows right here. There’s also Hindi in Agentforce Voice to reach beyond the English-speaking crowd.
Customers want AI but not at the expense of their data. We’ve made sure of that. It’s part of a wider shift in our products, from sales to analytics, where you don’t just have a static board in front of you; you can see why you should be acting on, say, the two in ten opportunities that are most likely to close.
Workforce and footprint growth
The numbers tell the story of our bet on this market. I came in when we had 2,500 in India; now we’re over 17,000. We’ve gone from four cities to seven, with a new tower in the works in Bengaluru.
Agentic AI as a CX differentiator
There’s a reason enterprises are keen on agentic AI. It can do the work within its lane and trigger things across your systems. Put it in a call centre and you might have 10 people creating a 10-minute hold. With agents, you can cut that to 10 seconds and not have to put 100 more heads on payroll.
Of course, you need guardrails. An agent should know when to let a human in. When you get that balance right, you’re looking at a 10-minute wait time that is now zero. That’s good for your NPS and your bottom line.
Healthcare is a prime example. Let the agent find the specialist and sort out the calendar. The patient gets a slot and it’s done. It doesn’t replace the doctor, it just frees them from the tedium of coordination.
Here are the India-facing moves Bhattacharya highlighted:
– Hindi support for Agentforce Voice for non-English markets.
– MuleSoft on Hyperforce in India for local hosting.
– AI infused across sales, service, marketing, commerce, analytics.
– Workforce scaled from 2,500 to 17,000-plus.
– Presence expanded from four to seven cities.
Cloud hesitation and the adoption divide
Still, some, in the public sector especially, are wary of the cloud. I’d argue it’s no less secure than what you have on-premises. Most of your clients aren’t in the security business, we are. It’s about logical as much as physical protection.
Then there’s the cost of it all. Gen AI and the like are hungry for compute. Trying to build that in-house is a barrier for many. Local hosting on something like Hyperforce is a way to have your cake and eat it too – you meet residency rules without being stuck with a weak setup.
You see a divide in how fast people are moving. The small and new ones are on it. The big, old-school organisations take their time because you have to retool the way you work. In that respect, some in India are behind what you see in the West, even if the startups are on a par.
What it means for healthcare players
BFSI is usually the first to the table and we’re seeing that in India. Now it’s trickling into life sciences, real estate and the like. Even in utilities and manufacturing there’s a steady appetite.
For pharma and providers, this is about what you can do today. Scheduling, triage, claims – agentic AI is built for it. And with multilingual and local options, there’s less to stand in the way of going live.
So it’s a matter of who gets in early. If you have the right mix of AI and human oversight, you’ll have an edge. As the regulatory side of things settles and people get more at ease with the cloud, I think we’ll see a lot more of it, and we have the talent in India to make it happen.
