The New Colleague: How Agentic AI Changes Enterprise Workflows

TL;DR –The New Colleague: How Agentic AI Changes Enterprise Workflows

• AI as a collaborator, not a tool:  Enterprises are embedding AI agents directly into workflows, treating them as digital colleagues that enhance capacity and consistency.

• Operational impact across industries:  From reducing administrative bottlenecks in healthcare to scaling wealth management services in banking, agentic AI is proving its value in domain-specific applications.

• The people challenge: Successful adoption depends on shifting employees from “tech-savvy” to “AI-savvy,” with a focus on trust, usability, and organisational culture.

• Balance as the differentiator: AI delivers efficiency, speed, and scale; humans bring creativity, empathy, and strategic judgment. Enterprises that align both will see the strongest returns.

The New Colleague: How Agentic AI Changes Enterprise Workflows

The rise of agentic AI signals a break from earlier forms of artificial intelligence. Where traditional systems are confined by rules and parameters, agentic AI can act with greater autonomy, pursuing long-term goals, making decisions, and mimicking human behaviour with little oversight. Its potential is sweeping, but so are the questions it raises: can people adapt to machines that no longer wait for instructions, and what does creativity look like when efficiency is increasingly automated?

 AIBP, together with Adobe, convened a private workshop, bringing together  Filipino enterprises during the 52nd AIBP Conference & Exhibition in Manila. Participants traced their AI journeys from tentative pilots to more ambitious deployments while confronting shared challenges. Industry-specific case studies served as a reminder that  it is no longer sufficient to keep a human “in the loop.” Sometimes, progress requires a partnership where the machine is not merely a tool, but a collaborator that makes work less painful and, perhaps, more humane.

Paperwork without Paper

At one of the Philippines’ largest hospitals, the arithmetic of healthcare is stark. 

An industry leader from Philippines' healthcare mega app, shared that the country has one doctor for every 26,000 patients, more than twice the recommended ratio. In rural provinces, the number stretches to one for 40,000, as most physicians gravitate to cities. Agentic AI is beginning to fill these gaps by removing administrative bottlenecks. 

“It depends on how fast the model responds. And that frees a doctor to really focus on the human connection which is, after all, why most doctors entered the field in the first place: to help people. It helps make doctors more efficient, less fatigued, and ultimately able to make better decisions. That translates into better outcomes for patients.”, shared Albert Padin, Chief Product Officer, at Metro Pacific Health Tech, MWell. 

Instead of a physician’s day being swallowed by prescriptions, reports, certificates, and endless paperwork, he shared that AI tools now listen in on consultations, transcribe conversations, and draft medical records in real time. This is especially helpful when there is a mountain of medical history where, instead of leafing through 50 pages of notes from previous visits, AI can distill a patient’s story into a concise summary, surfacing trends and anomalies that might otherwise be missed.

The problem of time and information overload isn’t unique to healthcare. Across industries, employees are buried in documents, data, and tribal knowledge that often sits locked away in silos. 

Adobe’s Ming Fai Chak observed the same principles apply across industries:

“A pragmatic step is training AI on internal resources: manuals, guides, policy documents so employees can quickly find what they need. That reduces dependency on colleagues and cuts time lost in information silos.”

When AI reduces friction in knowledge access, enterprises improve productivity, accelerate decision-making, and enhance employee satisfaction.

A Bridge for Bankers: Navigating Customer Expectations

In financial services, the challenge is not a lack of technology, but the diversity of customer behaviour. Many banks face a split market: digital natives demand intuitive, app-based services, while others remain tied to branch visits and personal interactions.

A technology head at one of the Philippines’ largest banks described the dilemma: over-investing in digital technologies risks alienating traditional customers, while over-reliance on branches risks losing ground to fintech competitors.

Agentic AI is emerging as a middle ground. For routine interactions, AI nudges customers toward digital channels while preserving the option of branch-based support. In wealth management, where relationships remain central, AI helps relationship managers prepare by generating personalised reports, analyses, and recommendations, tailored to a client’s unique profile. in real time..

Adobe’s Ming Fai Chak, Director of Solution Consulting, Asia, calls this shift a natural progression. The first stage is “human-augmented AI”—tools that support decisions and accelerate execution. The next is autonomy where the systems act as agents: drafting personalised emails or generating unique webpages automatically.

“Personalisation will be the next step. Once everyone gains confidence, you’ll want more things to be autonomous. And so we’re preparing for that future, where parts of our software operate as autonomous agents. Imagine an AI that drafts a personalised email, or instantly generates a webpage unique to one customer.”

The analogy with physical branches is direct. Just as a manager might lean on an app to frame a conversation, AI can create personalised experiences without inflating costs, a capability that will become critical as margins tighten and competition intensifies.

From Tech-Savvy to AI-Savvy

The harder task for Agentic AI is persuading people to adapt, not installing the systems themselves.

“Tech is always psychology.  It's always human behavior and adaptation rather than the tech part. So I hope we put a balance on that,” said Donald Lim, COO at DITO and Udenna. 

GCash, the Philippines’ leading e-wallet and finance super-app, demonstrates what this looks like in practice. Beyond deploying AI assistants for every employee, GCash has formally integrated more than 50 robotic process automation bots into its operations. Each bot has a registered ID and reporting line, functioning as if it were another member of staff.

The rationale is scale and consistency. Routine, rules-based tasks can be offloaded to digital agents, allowing employees to devote more time to complex or customer-facing work. It also helps normalise change: when every workflow is paired with an AI agent, the debate shifts from whether to adopt AI to how best to use it.

“Personalisation and scale: you can’t achieve that with people alone. You need AI to make sense of it. Speed and scale are things AI can deliver. Another strength is in providing persistence, not just repeatability”

From Pilots to Partnerships

Enterprises are moving from pilots into structured deployment, embedding agents directly into processes that were once the sole domain of human staff.

What will differentiate leaders is the ability to integrate it into workflows, manage cultural adoption, and identify areas where autonomy creates measurable value. Agentic AI is not a “destination” tool but an operating model shift. Enterprises that treat it as a partner, rather than a pilot, will define the competitive benchmarks of the next decade.