US-based CRM company Salesforce’s bet on developing an AI-first platform marks a significant strategic shift in recent years. The company has a significant presence in India with over 13,500 employees across multiple cities, and apart from developing solutions for key sectors in India, there is increasing adoption of AI agents within the company.
India is Salesforce’s second-largest market, with Salesforce India reporting revenue of Rs 13,385 crore for the fiscal year ended March 2025, up 47% year-on-year. Salesforce’s major customers in India include Tata Motors, Tata Power, Dr Reddy’s, razor payamong others. On a global level, Salesforce is projecting lower-end revenue of $41.3 billion for fiscal year 2026.
We are focused on growing our customer base in the Agentic AI space with our suite. agent force 360, Salesforce has multiple proofs of concept (PoCs) with companies and enterprises in India, but production deployment has been slow.
“We may be late to the market, but we’re not going to bring a product to market unless we address the issue of trust and until the data remains, no matter who owns that data,” Arundhati Bhattacharya, president and CEO of Salesforce South Asia, said at a recent event in Delhi.
In an interview with Forbes India, Bhattacharya said that Indian companies are more cautious about implementing AI on a large scale than in the United States and other countries. She added that the quality of data for training Agentic AI: PoC Customer satisfaction improves, and speed to production and deployment increases.
Excerpt from the interview:
Q. Salesforce is known to be slow. capitalize Regarding the recent surge in AI. Please tell me how agent force Has the 2024 launch plan been implemented by customers?
I think what we do really well is implement agent force In your workflow. For example, automating workflows allows digital agents to automatically trigger routine tasks. This is where we excel. There are multiple CRM data organization It also enables how you respond to specific customers and intelligently use AI agents to improve interactions over time. We have always created experiences, so this is where we add value. We are in the engagement layer, not the transaction layer. And the engagement layer lets you see how people are responding to the technology.
Q. How does the rate of AI adoption in Indian companies compare to US and Western companies? What explains this trend?
In the West, acceptance of AI is increasing. We (India) are generally hesitant to take big steps and prefer to start with small scale deployments. I will also try to see how I can do it optimize Some people try to do it themselves because it is expensive. The problem now is that technology is evolving at a pace that makes it really difficult to keep track of its updates.
So even if you can create something at first, it’s not that easy to keep improving. In that regard, I feel Indian industry is willing to experiment and build some solutions themselves. In terms of whether we’re on par with Western companies that are able to implement this quickly, I’d say we’re still doing an average routine where we’re a little bit behind the rest. But ordinary people are getting used to AI on an individual basis, and digitally enabled people are using AI in every possible way. On the corporate side, there is still some hesitation.
Also read: Inside Arundhati Bhattacharya’s Salesforce India Handbook
Q. Could you tell us more about the characteristics of the Indian companies you are partnering with on a trial basis?
Most of our customers in India are Indian conglomerates and not multinationals. Multinational companies are experimenting with agent AI at their headquarters, not in India. Most of them are in PoC stage, but we expect more to move into production stage.
Common areas where we can provide agent AI include BFSI, support, etc. center,cars,contacts centerimprove customer satisfaction.
Q. What are the key challenges faced by Indian companies across the board in moving from PoC to production stage when it comes to AI implementation?
There are two types of challenges we hear about. One is about the data itself.—What do you think? tidy It’s the data and how quickly we were able to get there. The second is workflow automation. Implementation requires process changes. Two of the key challenges are:
Another area where people need to take the plunge is with very limited data when running a PoC or pilot. Limited data obviously doesn’t return the level of response that users want to get or use. So you need to encourage your customers to use better data to get satisfactory results.
Q. How did Salesforce bring AI to departments within the company?
We’re aiming for customer zero with our software, and we want to show that being an agent company allows us to spend more time with customers and shift our talent from sitting at a console to building real relationships with customers.
We use AI in a variety of areas, including training and answering questions. summarize Drafting emails and replies, creating sales plans, IT support, and basic code writing. We use AI to study sawdust, but we don’t have the opportunity to track it, so a small opportunity is lost.
Our support helpline, help.salesforce.com, previously staffed only by engineers, is now staffed by agents. Currently, approximately 2% of cases are still referred to manual agents. Approximately 77% of initial inquiries on the support helpline are resolved. That’s why we are implementing AI in a big way.
