Salesforce has put the Agentic Enterprise at the centre of its Dreamforce 2026 narrative, as the company brings together customers, technology leaders and innovators from across the world to discuss how AI agents are expected to change the way enterprises operate. Addressing the media during Dreamforce, Salesforce leaders Arundhati Bhattacharya, President and CEO, Salesforce South Asia and Gavin Barfield, VP, Solution Engineering, Salesforce South Asia spoke about the company’s focus on bringing humans and AI agents together, while highlighting the role of trusted data, existing enterprise investments and integration in enabling AI adoption at scale. With more than 50,000 attendees from over 150 countries and more than 2,000 sessions, this year’s Dreamforce is focused on how trusted cloud applications, AI agents and AI-powered interfaces are transforming enterprise workflows.
Salesforce’s 24 Years of Innovation
Salesforce’s Dreamforce narrative this year reflects the company’s evolution from CRM towards a broader enterprise technology platform that brings together applications, data, analytics and AI. The focus is now shifting towards what Salesforce describes as the Agentic Enterprise, where AI agents become part of everyday business processes and work alongside employees rather than operating as standalone technology tools.
For enterprises, this shift also brings a practical challenge: how to introduce AI into existing technology environments without disrupting investments already made in applications and infrastructure. That challenge is particularly relevant in markets such as India, where enterprises across industries operate a mix of modern cloud platforms and legacy technology. For Salesforce, the mandate is to bring humans and AI agents together and enable enterprises to build, deploy and scale these capabilities across business functions. The company sees trusted data as a fundamental requirement for this transition.
Focus for Salesforce this Year
The focus at Dreamforce 2026 is on moving beyond AI experimentation towards enterprise-wide deployment. For South Asia, Salesforce sees an opportunity to move from AI experimentation to the Agentic Enterprise, where humans and AI agents work together to drive growth, productivity and customer outcomes. The company points to the region’s scale, digital ambitions and technology talent base as important advantages in building and deploying agentic AI across industries. According to the information shared by PIB, India has 2.5 times the global average in AI-skill penetration, while 87% of Indian enterprises are actively using AI. Across ASEAN, AI adoption among enterprises increased from 4.3% in 2023 to 23.5% in 2025, according to the figures shared by ASEAN Secretary General. The opportunity, Salesforce said, is not limited to large enterprises. Developers, startups, technology partners and AI-skilled talent are also expected to play a role in building the broader agentic AI ecosystem.
What Does This Mean for South Asia?
Salesforce sees South Asia as a region with different levels of AI maturity but a common opportunity to use AI to change traditional ways of working and deliver more personalised customer experiences. India, Singapore, the Philippines, Indonesia, Thailand and Sri Lanka are at different stages of adoption. However, the diversity of these markets also creates opportunities for enterprises to develop use cases suited to specific industries and customer requirements. India’s large enterprise base, growing AI capabilities and technology talent are being positioned as key factors that can support the scaling of agentic AI.
Also read: Salesforce Takes Enterprise AI Beyond the Traditional UI at Dreamforce 2026
Salesforce has also expanded its startup programme across India, Malaysia, the Philippines, Singapore and Sri Lanka, signalling a broader focus on the regional technology ecosystem. The company cited the IndiaAI Mission, with an investment of Rs 71,037.2 crore and more than 38,000 GPUs, as another indicator of the country’s growing AI infrastructure and ambitions. Singapore, meanwhile, has committed more than S$1 billion in government investment towards AI research, talent and applied AI through 2030, according to the information shared by the Singapore government.
India and Sri Lanka are emerging as important markets in Salesforce’s South Asia strategy. In India, Salesforce highlighted its growing ecosystem, including more than 1 million learners to be skilled by 2030, the availability of Agentforce Voice in Hindi, and the Salesforce Tower Bengaluru, which the company described as its first such facility in India. Salesforce also pointed to the country’s growing Trailblazer community, with 3 million active Trailblazers, and the establishment of its India Centre of Excellence. Sri Lanka, meanwhile, is part of Salesforce’s expanding startup programme across the region. The broader focus is on creating an ecosystem that brings together enterprises, developers, startups and partners as AI adoption moves towards larger-scale deployment.
How Salesforce is Addressing the Legacy Infrastructure Barrier to India’s Agentic Enterprise Ambitions
In India, there are a significant number of AI pilots moving into deployment, but the percentage that progresses to large-scale deployment remains relatively low, according to various surveys. Legacy infrastructure is one of the challenges. Autonomous agents are often layered onto infrastructure that was not originally designed for the speed or level of autonomy that AI requires. When asked about how Salesforce’s new product line addresses the reality of legacy infrastructure, and if enterprises can realise the value of these technologies without having to undertake large-scale technology replacements, Salesforce leaders had the following to say to Tech Achieve Media.
Gavin Barfield: Legacy technology will exist, and does exist, in every organisation. I have never had the pleasure of going to too many organisations that do not have some form of legacy technology. The important first step is the consolidation of data. Most companies have data that exists across multiple systems, and that data is often inconsistent across those systems. AI will not work very effectively if you have poor-quality or fragmented data across multiple systems. So, the first step is to determine how you consolidate that data. That is why Data 360 and the architecture I showed you earlier are so important. You need to create that layer of trusted context.
This involves moving or referencing data across your legacy systems. The reason why we acquired Informatica earlier this year was precisely to address this need. You need to be able to bring together data sources from legacy applications. We already have MuleSoft as part of our portfolio, which is able to do this. Informatica, however, provides the trust and data governance required to look at data across legacy applications, define your data dictionary, establish your source of truth, and provide governance and traceability. This helps create a solid data platform through Data 360, which is the foundation and fundamental building block for AI.
So, you will never completely get away from legacy technology. It will continue to exist. Most companies, I don’t think any company, will be able to put their hand on their heart and say, “I’ve got clean data.” Every company of any scale will have data inconsistencies and fragmented or messy data across multiple systems. But with the Data 360 layer, we are helping, through products such as Informatica, Tableau and MuleSoft, to bring that layer together and create a single source of truth, which becomes the basis for AI to run on.
Arundhati Bhattacharya: One of the things I would also like to supplement here is that we are not very much in favour of a total rip-and-replace approach. We try to ensure that we utilise whatever investments the company has already made, whether in a data layer or in other applications, and bring those into the particular layer alongside our offerings to deliver the best results. That is one of the reasons why our Data 360 solution has a zero-copy architecture, so that the data is not replicated. If the data is already residing in a data lakehouse, we are able to bring the intelligence to Data 360 without copying the data and creating unnecessary replication. This has all been thought through to ensure that whatever legacy pieces an organisation has can be utilised to the greatest extent possible while implementing an agentic enterprise.
