From Reactive to Proactive: How Balibago Waterworks Is Building an AI Digital Twin

Balibago Waterworks is the fourth-largest private water utility in the Philippines, operating 93 municipal franchises across 24 provinces and serving 1.5 million people through public-private partnerships with local governments. For most of its history, maintenance followed a simple pattern: something breaks, someone fixes it. President Criselle Alejandro set out to change that.

The Problem: Manual, Paper-Based, Reactive

Before its digital transformation, Balibago had no systematized approach to preventive maintenance or asset management. Everything ran on paper and reacted to emergencies as they happened. Alejandro wanted to shift the company from reactive to proactive, and to close the gap between Balibago and utilities already using digital tools.

Building Balibot

The company partnered with TeamSolve, a connection made through Imagine H2O, to pilot a digital twin now known internally as Balibot. The first version ran through WhatsApp as a chatbot, but an early language mismatch (chats in English, operators speaking Filipino) meant it needed to be rebuilt in the local language before it could gain traction. TeamSolve then moved off WhatsApp entirely and developed a dedicated platform. Balibago’s team jokes that the pilot was “a bicycle” and the current version is “an SUV.”

Early Results and Adoption

Balibot launched across seven pilot sites, with an incentive program built in to encourage plumbers and pump operators to explore its full feature set. Ten months into use, the system tracks the condition of pumps, generator sets, and water treatment plants, and its agentic AI now generates maintenance recommendations for managers to act on.

Alejandro is direct that it is still early. The utility hasn’t yet quantified reductions in downtime or cost, and case studies demonstrating clear ROI are still a year or so away. What has changed is behavior: teams are now sent to intervene before failures occur, not after.

Advice for Utilities Starting This Journey

Alejandro’s clearest lesson for other utilities considering similar AI tools: the technology isn’t the hard part, the data is. Balibago’s original data collection produced messy, inconsistent insights, requiring the team to redesign its forms so information could be captured in a way both humans and AI could interpret. Training the AI, she notes, is not automatic. It takes deliberate, ongoing work.

Looking Ahead

Balibago’s experience is a useful data point for utilities serving smaller, dispersed communities who may assume digital transformation is only viable at scale. Two years into the process, Alejandro’s team is still building toward the case studies that will prove the model out, but the operational shift from reactive to preventive maintenance is already underway.