By Saurav Gupta
Agentic AI combines autonomous capabilities with human oversight to enhance productivity and value creation. By integrating real-time contextual data through smart data fabrics, organisations can enable AI agents to support decision-making while ensuring ethical governance. This hybrid approach empowers teams to focus on innovation, customer service, and strategic growth.
Providing more sophisticated, dynamic, and interactive capabilities than robotic process automation or virtual assistants, agentic AI has the potential to supercharge the efficiency of workflows, delivering clear productivity gains in day-to-day work. A fast-growing area, Gartner believes a third of enterprise software applications will include agentic AI by 2028 – up from almost zero last year (2024).
But while agentic AI in business represents advanced form of AI that operates autonomously, for perhaps the next five years the most effective implementation of agentic AI will retain human supervision at their core. This is largely because of lack of contextual understanding, immaturity of AI governance regulations and ethical oversight.
Varying degrees of human supervision
The extent of human involvement will vary according to use cases. In supply chain platforms, agentic AI can scan and analyse supplier data, shipping schedules, and compliance updates in real time. But a human will make the key decisions that reduce friction across operations.
In the investment and banking sector, agentic AI is set to process research reports and data at a pace far beyond any analyst’s capacity while applying human judgement to produce hyper-personalised summaries that match each user’s needs based on their risk profile.
What agentic AI enables is a shift for organisations to start looking at their workforce as value creators.. Businesses will free up time previously spent on repetitive analysis, enabling teams to focus on decisions that drive value, improve their ability to serve customers, and free up time for greater concentration on innovation and growth.
Human-Agent collaboration
The most successful organisations of future will develop a culture of human agent collaboration that keeps human oversight at the heart of agentic AI enabled systems. This requires organisations must upskill their workforce to develop hybrid skills that combines their expertise with AI literacy.
We are set to see a significant move away from clicking on screens and requesting dashboards when using AI agents, to natural language interactions using everyday speech or text. Chatbots will become more advanced in handling these queries even as they become more complex. Advances in prompt-engineering mean enterprise-level agentic AI networks will adapt themselves to workflows in each organisation
Enterprises must implement systems with AI agents that need to operate within safe and ethical boundaries aligning with organisational policies and compliance with regulations. Humans are key to providing ethical oversight, upholding organisational standards and managing risks in agentic AI enabled systems.
Creating the right foundations
To unlock the full benefits of agentic AI, organisations will need to address data management and the architecture that supports it. The large language models (LLMs), on which agents depend, remain susceptible to hallucinations – inventing facts or applying information erroneously. Contextual data is required because LLM data can be a year old. If for example, an agentic AI application is to provide a business user with a live view of a customer, it must not present stale data – or hallucinatory information. It needs real-time contextual data to provide an accurate, live view. The real power lies in multiple agents that can communicate and coordinate with each other on a connected, real-time and trusted data infrastructure.
Organisations must bring together data from these multiple sources in a trusted way, applying sturdy guardrails. Audit trails are essential to ensure data is secure, accurate, and to ensure it is used responsibly. Data must be current, real-time, transparent and auditable in an observable, explainable AI model.
Smart data fabric architecture and agentic AI
The persistently siloed nature of enterprise data remains a significant hurdle. Data within an organisation is of varying types – structured, semi-structured and unstructured. Organisations need to harmonise the data from all sources, and employ effective data governance, so that when they use LLMs and contextual data, the output is what is required.
This is where data fabric comes in which acts as a smart data layer that connects and manages data from all your systems in real time. It eliminates data fragmentation by seamlessly integrating every source, ensuring consistency and accessibility of data. Data fabrics utilise metadata management , knowledge graphs and semantic layers to add context and meaning to data. This enables AI Agents to understand business context and relationships between different data points. This fulfils the basic needs of agentic AI –leverage unified data to fuel AI models that deliver accurate, context-aware insights for decision making and task automation.
Data fabric using centralised data architecture and governance model, allows data to be shared and integrated across the entire organisation. The alternative to a fabric is a more decentralised infrastructure that makes agentic AI difficult. Agentic AI systems require multiple agents which is federated by definition but having federated data infrastructure and governance adds another layer of complexity and even more coordination across people and processes.
Building the right data architecture is critical to agentic AI. But once these firm foundations are in place, businesses can roll out agentic AI – not to replace people but to amplify their capabilities. Agentic AI will help them move faster, think bigger, and will keep human judgment at the heart of decisions that matter.
Businesses will be able to liberate their teams from routine clerical and admin tasks or repetitive analysis, enabling them to focus on innovation and growth.
Sales Engineer, InterSystems 
