Enabling AI-Driven Network Intelligence and Investment Optimization for a Leading Southeast Asian Telecommunications Provider
AI-powered Digital Brain delivered 50% churn reduction opportunity, THB 360M+ revenue uplift, and 60% faster network investment decisions through predictive analytics and intelligent prioritization.
Built on Amazon SageMaker and Amazon Bedrock
50%
churn reduction opportunity identified
THB 360M+
projected revenue uplift via optimized investments
~60%
reduction in network investment decision cycles
Overview
About the Customer
The customer is a leading telecommunications operator in Southeast Asia, serving tens of millions of mobile subscribers across urban and rural markets. The organization delivers nationwide voice, data, and digital services and operates one of the region's largest network infrastructures.
Operating in a highly competitive market, the company focuses on improving customer experience, optimizing capital investments, and strengthening long-term profitability through data-driven operations.
The Challenge
Business Challenges
- The operator was experiencing rising customer churn, lower net-adds and declining satisfaction driven by inconsistent network performance and delayed issue resolution. Although large volumes of network, customer, and complaint data were available, this information was fragmented across multiple operational and analytical systems, limiting its usefulness for decision-making.
- Network investment planning and fault prioritization were largely manual and reactive. Decisions on site upgrades, capacity expansion, and service remediation relied heavily on expert judgment and ad hoc reporting. This resulted in long analytical cycles, inconsistent prioritization, and increased risk of inefficient capital allocation.
- In addition, regional and provincial teams lacked timely access to actionable insights. Limited visibility into root causes of service degradation made it difficult to address performance issues before they impacted customers.
- Without a systematic, AI-driven approach, the operator faced continued churn, inefficient network investments, and reduced ability to compete in a market dominated by a small number of major players.
The Solution
Solution Approach
The operator engaged OneByZero (OBZ) to design and implement an AI-powered 'Digital Brain' to enable predictive network analytics, root cause analysis, and intelligent investment prioritization on AWS.
Unified Data and Analytics Foundation
OBZ established a centralized data platform to integrate network, customer, revenue, and complaint data into a unified analytical environment. Internal operational datasets were combined with external performance and topology data to create a single source of truth for network intelligence.
This foundation enabled consistent access to trusted data across engineering, operations, and business teams.
Machine Learning and Explainable AI Framework
OBZ implemented a production-grade machine learning framework on Amazon SageMaker to enable predictive churn analytics and structured root cause analysis across network and customer domains. Supervised models were trained using a high-dimensional feature space combining network KPIs, ticket and complaint data, subscriber behavior, and revenue attributes. Beyond predictive modeling, the framework incorporated structured driver attribution to identify which network degradations and operational issues were most strongly linked to churn and revenue impact. To ensure transparency and adoption, explainable AI techniques (SHAP) were embedded into model outputs. Models were deployed with version control, validation, and monitoring to ensure production reliability.
Generative AI–Driven Decision Support
OBZ implemented a generative AI layer using Amazon Bedrock to enable natural language interaction with analytical insights. Operational and regional teams could query network performance, churn drivers, and site-level recommendations through conversational interfaces, reducing reliance on manual reporting. Amazon QuickSight dashboards provided real-time visibility into network health, churn risk, and prioritized remediation opportunities across regions. Together, conversational AI and interactive dashboards enabled faster insight retrieval, improved cross-functional alignment, and more data-driven decision-making.
End-to-End Governance and Deployment
The solution was deployed in a secure, multi-account AWS environment aligned with telecom governance and compliance requirements. OBZ provided end-to-end delivery and support, including solution architecture and security design, data platform and ML pipeline deployment, model development, validation, and tuning, production deployment and monitoring, and governance and access controls. This ensured production reliability, scalability, and regulatory alignment throughout the implementation.
AWS Services Used
Built on AWS
Amazon S3: Centralized data lake for raw, curated (silver/gold), and model-ready datasets, enabling cost-effective and scalable storage across network, customer, and revenue domains.
AWS Glue: Automated data ingestion, ETL pipelines, data transformation, and centralized metadata management through AWS Glue Data Catalog.
Amazon Athena: Serverless interactive query engine for ad hoc analysis, data validation, and rapid exploratory analytics across the data lake.
Amazon Redshift: High-performance, distributed data warehouse for complex analytical workloads and executive reporting.
Amazon SageMaker: End-to-end machine learning platform for feature engineering, model training, hyperparameter tuning, batch inference and monitoring of predictive and diagnostic models.
Amazon Bedrock: Generative AI foundation layer enabling natural language interaction, conversational analytics, and AI-assisted insight retrieval across structured and unstructured datasets.
Outcome
Business Outcomes
- Enabled frontline and regional teams to make faster, data-backed decisions through conversational AI and real-time dashboards
- Identified network segments with up to 50% churn reduction opportunity through impact-based prioritization and enabled THB 360M+ projected revenue uplift via optimized investments
- Reduced network investment decision cycles by approximately 60% by replacing manual analysis with automated intelligence
- Improved prioritization of high-impact sites, helping reduce inefficient capital allocation
- Standardized root cause analysis across regions, improving consistency and governance
- Increased organizational adoption of analytics through self-service access to trusted insights
- Overall, the solution shifted network management from a reactive, report-driven process to a predictive, intelligence-led operating model.
About OneByZero
OneByZero is a frontier Systems Integrator specialising in AI Coworker design, build, and deployment for regulated enterprises across financial services, telecommunications, and retail. We operate across ASEAN, India, ANZ, and Japan, with a presence in the United States, combining domain expertise in regulated industries with deep AI engineering capability.
HQ Singapore · Tech HQ USA · Australia · India · Indonesia · Malaysia · Philippines · Thailand · Vietnam
