Enterprise Artificial Intelligence: Transforming Workflows with Agentic Systems

Enterprise artificial intelligence is evolving beyond chatbots into autonomous systems known as agentic systems, capable of handling complex tasks with minimal human intervention. This shift is beginning to transform hiring, workflows, and decision-making processes within organisations.

Salesforce’s AI Transformation

This transformation is evident within Salesforce itself. CEO Marc Benioff has highlighted how AI-driven coding tools are enhancing productivity, leading to reduced hiring needs.

Benioff noted that there will not be an increase in the number of engineers hired for FY26, as the productivity from coding agents has provided the necessary capacity for the year.

These statements reflect a greater transition occurring within enterprises: AI is no longer merely supporting employees; it is actively performing the work.

Enterprise General Intelligence (EGI) at Salesforce

For Salesforce, this development signifies the initiation of what is referred to as Enterprise General Intelligence (EGI). In a discussion with Startup Superb, Deepak Pargaonkar, Vice President of Solution Engineering at Salesforce India, shared that EGI is not a singular model but a comprehensive capability intended to operationalise AI throughout business workflows.

Pargaonkar explained that it necessitates an operating system that supplies agents with governed data and context, encodes existing business logic, and offers visibility into the activities and rationale of the agents.

The Agentic Enterprise Stack

Salesforce’s strategy centres around the “agentic enterprise” framework, comprised of four layers: context, work, agency, and engagement. Pargaonkar elucidated that the aim is to transform raw intelligence into tangible work and progress from individual agency to enterprise agency, coordinating thousands of agents and humans to manage intricate tasks across various teams, functions, and companies.

From Pilots to Large-Scale Implementation

The transition from testing to full-scale deployment is already progressing. Salesforce’s CIO study revealed a staggering 282% rise in AI implementation during 2025, indicating that businesses are increasingly integrating AI into their core operations rather than treating it as a side project.

India is leading this shift, with Salesforce’s State of Sales data showing that 91% of Indian sales professionals consider AI agents vital for business success.

Pargaonkar remarked that these are no longer mere experiments but are enhancing human judgement in actual workflows.

Challenges of Remaining in Pilot Mode

However, companies that continue to experiment often face fundamental challenges, particularly concerning data readiness and governance.

Pargaonkar observed that the intelligence of AI agents depends on the unified, real-time information they access. Leaders who recognise this are moving at a fast pace.

Sector-Specific Adoption Trends

Adoption rates are not uniform across all sectors. Financial services, healthcare, and manufacturing are at the forefront in adopting agentic systems, largely due to their operational structures.

Pargaonkar mentioned that financial services firms possess robust data infrastructures and regulatory requirements that align well with agent-driven automation, making deployment a natural extension of their existing practices.

In healthcare, the high volume of administrative tasks, including documentation and coordination, presents an opportunity for agent technology to excel.

The Deployment Challenge

Despite the growing enthusiasm, scaling enterprise AI remains complex. Customisation remains a significant challenge.

Salesforce is addressing this with its Agentforce ecosystem, which includes prebuilt agents and templates tailored for specific industries. The AgentExchange marketplace currently offers nearly 800 reusable agent assets from over 160 partners to expedite deployment.

Nevertheless, Pargaonkar emphasised that customisation is unavoidable and crucial. The focus for customers should be on areas that yield the most value, such as providing agents with appropriate business knowledge, configuring them for particular workflows, and verifying data readiness.

He asserted that data quality is the key determinant of performance, stating that a well-deployed agent operating on clean, governed data will consistently outperform a sophisticated agent relying on fragmented data.

Redesigning the Talent Landscape

As enterprises expand their use of agentic systems, the implications for employment become increasingly evident. Benioff’s comments on hiring reflect a broader realignment occurring in the technology sector.

Pargaonkar believes this transformation presents more opportunities than it takes away. He stated that the talent pyramid is evolving, and the shift creates more opportunities rather than simply displacing existing roles.

The real challenge lies in how organisations reinvent their work processes. Employees will need to acquire new skills to manage and collaborate with AI agents while their focus shifts towards higher-value functions.

Pargaonkar highlighted that India has a remarkable talent pool and the challenge is to redirect that talent towards skills that agents cannot replicate, such as judgment, creativity, and the ability to facilitate large-scale human-AI collaboration.

As enterprise AI transitions from simple chat interfaces to autonomous execution, the organisations that thrive may not be those with the most advanced models, but those that can seamlessly integrate agents into their workflows.