PLDT Reports Faster Ops From Three Agentic AI Systems
Philippine telecommunications group PLDT reported operational gains from three UiPath-centred agentic systems—KAI, Ellie, and ERICA—in announcements around September 10–11, 2026. KAI targets knowledge retrieval for proposals and support; Ellie connects enterprise enablement workflows; ERICA supports annual risk reviews with scoring and treatment suggestions. The company previously introduced the tools separately; the latest package brings them together with time-saved estimates.
Filed under Industry and dated September 11, 2026, this AI4Philippines briefing treats PLDT’s productivity claims as company estimates, not independently audited metrics. Reported ranges include tens of thousands of hours saved yearly for KAI and Ellie, sharply shorter risk-evaluation windows for ERICA, and migration stories from earlier Copilot-based prototypes onto UiPath with Teams integration. Governance notes stress grounding KAI in approved sources—useful, but incomplete without published access-control detail.
Why it matters: BPO and shared-services hubs in Metro Manila watch PLDT as a large domestic reference architecture. Concrete agent deployments matter more for Filipino operators than abstract megawatt masterplans already covered elsewhere, because they show how telco-scale knowledge work actually changes when retrieval and risk agents sit inside governed automation platforms.
What it means in practice
Filipino operators should pick one workflow with measurable delay or error; confirm lawful data access; assign a human owner; run a time-boxed pilot; and publish honest metrics including failure cases. Prefer stacks that log prompts, sources, and overrides. Align with Philippine privacy and labour expectations. Do not copy headline hour-savings without matching measurement methods.
Workers on enterprise sales and risk teams deserve clarity when assistants change performance targets. Caveats: overlapping hour estimates across tools can double-count benefits, and broad percentage reductions often hide variance by case type. AI4Philippines will keep asking for methods, not only marketing ranges.
What to watch next: whether PLDT publishes audited KPI methodologies; expansion beyond internal teams into customer-facing channels; and peer telco or BPO replications using similar agent patterns. Readers can continue on the AI4Philippines homepage for related stories, or browse the Newsroom for additional briefings.
Bottom line: treat this update as orientation, not instruction. Filipino enterprise agent deployments are real, uneven, and still early. Organizations that benefit most will test tools on Filipino problems they already understand, measure honestly, and keep people responsible for outcomes.