AI Agents: Redefining Business Transformation

IA FORUM MEMBER INSIGHTS: THOUGHT LEADERSHIP INTERVIEW

Santhosh Vijayabaskar, Director, Intelligent Automation & Process Excellence, FIDELITY INVESTMENTS

How are AI agents changing the landscape of business transformation?

AI agents are fundamentally rewriting the rules of business transformation by shifting from static automation to dynamic, cognitive collaboration. These agents are not merely tools but decision-makers that can independently execute complex workflows, collaborate with other agents, and adapt dynamically to changes. Unlike traditional tools that follow rigid scripts, agents - powered by advancements like the new OpenAI’s Operator framework or the latest DeepSeek’s reasoning-optimized LLMs - act as strategic partners.

In practical terms, this means our businesses can now scale decision-making without human bottlenecks, allowing us to focus on strategy rather than repetitive problem-solving. Imagine a supply chain that self-heals during disruptions or a marketing team where AI agents A/B test campaigns in real time while you sleep. This isn’t just efficiency gains - it’s about creating adaptive organizations that learn and evolve at the speed of market change. The key shift here is that AI agents are no longer just assistants - they are collaborators, fundamentally transforming how our businesses innovate and operate.

What makes AI agents particularly effective in driving process excellence?

They’re the Swiss Army knives of efficiency. Traditional automation is like a metronome - steady but rigid. Unlike traditional automation tools that follow predefined scripts, AI agents leverage capabilities like natural language processing, reinforcement learning, and data-driven insights to adapt

to real-time scenarios.

For example, in financial services, we can transform fraud detection: instead of rigid rule sets, agents can now analyze transaction patterns, customer behavior, and global risk signals to flag anomalies with surgical precision. The magic is in their contextual intelligence - they see patterns we miss and turn “good enough” into “exceptional”. Multi-agent collaboration systems also play a pivotal role, enabling seamless handoffs between tasks and departments, thereby eliminating silos. Ultimately, their adaptability and capacity for learning make them catalysts for continuous improvement.

While this tech is shiny, let’s ground this in reality: AI Agents are surely not a silver bullet when your process is broken. Agents excel when paired with disciplined process design and clean data pipelines. They’re not replacements for operational rigor; they’re accelerators of it.

How do you ensure AI agents align with broader business goals and strategies?

Alignment starts with clarity of purpose. AI agents must be deployed to solve specific business problems, not as blanket solutions looking for relevance. It’s crucial to integrate them into the organization’s strategic roadmap, ensuring their outputs directly support broader goals like customer satisfaction, operational efficiency, or revenue growth. I advocate for “strategy layers“- where business KPIs are translated into agent-level objectives.

For example, if you are in a retail space you can map your “customer lifetime value” goal into specific agent behaviors: personalization engines prioritize high-potential shoppers, while inventory bots dynamically adjust stock based on predicted loyalty impacts. Governance frameworks are equally important. Platforms like AI ethics auditors and governance tools can monitor bias, ensure data security, maintain ethical boundaries to ensure agents don’t just perform tasks but internalize strategic priorities. It’s like teaching chess to AI: you want it to win (business outcomes), not just move pieces (complete tasks).

What challenges do companies face when integrating AI agents into their operations?

One of the most significant challenges is cultural resistance. The idea of AI agents making autonomous decisions can evoke fear or skepticism among employees. Building trust requires transparent communication about their role and the value they bring, alongside training programs to help employees adapt to hybrid work environments. Another challenge lies in the technical domain - specifically, ensuring data quality and availability. AI agents thrive on accurate, diverse, and real-time data, yet many organizations struggle with legacy systems or data silos. Moreover, security concerns loom large; many enterprises report vulnerabilities in AI systems. Addressing these challenges demands robust data pipelines, cybersecurity protocols, and cross-disciplinary expertise.

What role does intelligent automation play alongside AI agents in achieving process excellence?

I have to say, this is one of my favorite questions. Intelligent automation and AI agents are complementary forces in achieving process excellence. Intelligent automation focuses on predefined workflows and process optimization, while AI agents introduce a layer of autonomy and cognitive decision-making. Let’s consider invoice processing: RPA extracts data, but agents analyze payment terms, predict cash flow impacts, and even negotiate with supplier bots. Together, they create an ecosystem where routine tasks are automated, and more complex, strategic tasks are delegated to AI agents. The key is designing feedback loops - automation executes, agents learn, and processes evolve. It’s not about replacing humans; it’s about creating a collaborative triad: where humans set vision, agents strategize, and automation executes.

How do you measure the success of AI agents in business transformation initiatives?

Measuring AI success isn't about chasing vanity metrics like “tasks automated” or getting lost in a sea of data points. It's about capturing the full spectrum of impact, from cold, hard numbers to the subtler shifts in your organization's DNA. I've distilled this down to what I call the "3P Framework".

It's straightforward, adaptable, and - most importantly - it speaks the language of business, not just tech.

1. Performance: This is your technical foundation. We're talking accuracy, precision, and recall. But don't get bogged down in jargon. What matters is this: Is your AI making the right calls, consistently? In a financial services context, for instance, we're looking at how often our risk assessment agent correctly flags potential fraud without crying wolf.

2. Process: Here's where the rubber meets the road. How is your AI transforming operations? I'm looking for efficiency gains - hard numbers on time saved or productivity boosts. But I'm also keenly interested in adaptability. How quickly does your agent pivot when the market throws a curveball? A trading algorithm that can't adjust to sudden volatility isn't just underperforming; it's a liability.

3. People: This is the secret sauce, folks. The best AI in the world is useless if your team doesn't trust it or know how to leverage it. I want to see user satisfaction scores, adoption rates, and - crucially - tangible business impacts. Did your customer service AI actually move the needle on retention rates? That's the kind of metric that gets C-Suite attention.

These 3 aren't siloed categories. They're interconnected. An AI that performs well technically but disrupts established workflows will tank in the “People” category. Conversely, an agent that's embraced by your team but can't deliver consistent results is just a fancy paperweight. The beauty of this “3P” framework is its flexibility. Whether you're in fintech, healthcare, or manufacturing, these principles apply. It forces you to look beyond the algorithms and consider the holistic impact on your business.

What emerging trends in AI agents excite you most for the future of business operations?

Four trends keep me up at night (in the best way):

1. AI Agents with “Gut Instincts”: The Latest DeepSeek’s R1 model that weigh risks like a veteran CEO.

2. Physical-Digital Fusion: Vision-language models controlling warehouse robots, as seen in NVIDIA’s Project GR00T.

3. Agent Collectives: Swarms of specialized agents (e.g., pricing, compliance, CX) forming on-demand task forces.

4. Self-Policing Ecosystems: Agents that auto-redact sensitive data in real time, turning GDPR compliance from a checkbox into a reflex.

In summary, what are your key lessons learned and what advice would you share with others tackling similar objectives or challenges?

AI agents are powerful, but their value is only realized when they serve clear, strategic purposes aligned with broader business goals.

My advice is threefold:

1. Start small but think big. Pilot AI agents in specific areas, measure success, and scale strategically.

2. Focus on the human element. Engage employees early, invest in upskilling, and foster a culture that sees AI as an enabler, not a threat.

3. Embrace continuous learning. AI agents are evolving rapidly; staying informed about emerging trends is essential to maintain a competitive edge.

Above all, remain agile. The organizations that succeed in leveraging AI agents will be those that view transformation as an ongoing journey, not a one-time project.

Author Disclaimer: The views and opinions expressed herein are those of the Author alone and are shared in a personal capacity, in accordance with the Chatham House Rule. They do not reflect the official views or positions of the Author’s employer, organization, or any affiliated entity.