PLDT turns manual effort into intelligent action with agentic AI built on UiPath

AI agents created with Agent Builder in UiPath Studio, equipped with UiPath Robots as their tools and connected to platforms like Microsoft Outlook and SharePoint through UiPath Integration Services – save tens of thousands of hours annually and free teams for higher-value work.

With 5G adoption accelerating, broadband demand expanding, and satellite connectivity poised to grow in the Philippines, PLDT operates in a dynamic market where enterprise customers expect faster, more reliable service and support.

The 2024 PLDT Enterprise Net Promoter Score (NPS) survey reinforced the importance of these expectations, showing that customer confidence is shaped not only by network performance, but also by how quickly issues are addressed and how clearly customers are kept informed. The survey highlighted the value of responsive relationship managers and a consistent loop of feedback and updates throughout issue resolution.

To support that shift, PLDT turned to artificial intelligence (AI) tools that could integrate with existing systems and give employees faster access to reliable information. Since May 2025, the organization has partnered with UiPath to apply intelligent automation in ways that augment teams, improve responsiveness, and support customer experience priorities tied to NPS.

With agentic AI capabilities, we're not just automating tasks, we're embedding intelligence into our workflows.

Gilbert Gaw, PLDT and Smart First Vice President and Head of IT and Transformation Office at PLDT

Faster answers, stronger service quality

Sales, presales, and relationship teams often spent significant time sifting through emails, technical documents, and other repositories before they could answer customer or internal queries. This created additional lead time for proposals, escalations, and follow-ups at touchpoints that influence NPS, and pulled focus away from consultative work with customers. They needed a more efficient way to access the right information.

Enter Project KAI, which introduced KAI (Knowledge, Automation, and Intelligence) as an AI assistant for enterprise knowledge retrieval. Teams can ask for the information they need and receive fast, contextual responses grounded exclusively from approved sources. KAI helps surface product specifications, facility availability, service readiness documentation, and network performance data so employees can respond to customers and internal teams with greater consistency and speed.

Together with UiPath, the Intelligent Automation team at PLDT worked to create and refine KAI, introduce cybersecurity guardrails, and iteratively enhance response accuracy. KAI now handles 11,000 of 11,300 queries per month, demonstrating strong adoption at scale. Compared with the one-to-five-day turnaround previously required for many information requests, KAI returns answers in one-to-three seconds, enabling an estimated 99% faster response rate.

Beyond speed, KAI reduces approximately 25,000–30,000 workforce-hours of manual effort annually, equivalent to around 12 full-time (FTEs) of productivity capacity. This increases organizational capacity without additional headcount and allows teams to spend less time on repetitive information gathering and more time applying judgment, managing relationships, and delivering higher-value, consultative support to customers—particularly at touchpoints that influence NPS.

Better preparation for better relationships

Deep customer understanding, solution alignment, and rapid access to enterprise knowledge are critical across PLDT Enterprise—from sales and presales to product and support teams. Traditionally, these activities required significant manual effort across research, proposal creation, competitive analysis, and solution mapping, slowing execution and reducing operational efficiency. But with Ellie, this is all significantly reduced.

Ellie transforms this experience by serving as PLDT Enterprise's AI-powered business intelligence and productivity assistant—purpose-built for enterprise operations, customer engagement, and solution enablement. Unlike generic AI tools, Ellie is designed around PLDT Enterprise's products, internal knowledge, operational workflows, and customer requirements, enabling teams to access accurate, contextual, and enterprise-specific insights in real time.

Integrated into existing workflows and available 24/7, Ellie enables employees to retrieve information instantly, generate customer-ready outputs, and make faster, more informed decisions. From creating tailored proposals and presentation materials to delivering competitive comparisons and solution recommendations, Ellie significantly reduces preparation time while improving consistency and response quality.

Its solution comparison capability enables teams to evaluate offerings side-by-side across features, use cases, deployment models, pricing structures, service inclusions, and business value—helping improve recommendation accuracy, strengthen positioning, and accelerate customer engagements.

Beyond sales enablement, Ellie also functions as a centralized enterprise knowledge and operations hub. It supports FAQ resolution, advisory updates, pricing inquiries, and intelligent inventory management for enterprise devices and network assets, including mobile phones, tablets, pocket WiFis, SIM cards, routers, switches, and related infrastructure. This provides teams with improved visibility, utilization tracking, lifecycle monitoring, and operational accountability across distributed assets.

Powered by intelligent document analysis, contextual knowledge retrieval, automated content generation, and personalized recommendations, Ellie minimizes repetitive manual work and enhances productivity across teams. By enabling employees to focus on higher-value customer and strategic activities, Ellie helps PLDT Enterprise accelerate deal cycles, improve engagement quality, strengthen cross-functional collaboration, and scale operations more efficiently.

Derisking, and fast

At PLDT and its wireless communications and digital services subsidiary Smart Communications, Inc. (Smart), analysts used to spend weeks on annual risk cycles. The increasing frequency and depth of reviews meant that they needed risk intelligence that was faster, more consistent, and more scalable—without compromising governance standards. This led to the introduction of agentic AI system ERICA (Enterprise Risk Intelligence Companion Agent), which would help them cope with the speed and scale of the risk cycle.

ERICA's Risk Statement Generator helps by synthesizing organizational inputs into structured, policy-aligned risk statements, while its Risk Assessment Agent scores inherent and residual risks, labels and categorizes them, and recommends treatment options. Since deploying ERICA in February 2026, efforts for this part of the enterprise risk cycle have been reduced by 97–99%, and tasks that took 2 to 10 days can now be completed in five minutes to one day.

With training and enablement provided by UiPath, the team at PLDT and Smart have also trained ERICA to produce outputs that consistently meet enterprise risk management standards. Users can now shift their focus to strategic risk treatment rather than data synthesis. Smart has also increased its capacity to conduct timely, high-quality risk reviews.

Outcomes:

• Faster, more accurate work: AI reduces manual effort while accelerating turnaround times and improving accuracy

• More output without added headcount: automation increases team productivity and capacity without requiring additional headcount

• AI that scales across the enterprise: KAI, Ellie, and ERICA address targeted needs across individual users, teams, and organization-wide workflows

Additional resources:

Explore UiPath solutions for telecommunications. Discover how AI agents and automation can help telecommunications providers streamline operations, improve customer experiences, and deliver faster service.