Enabling AI-Driven Customer Engagement and Churn Mitigation for a Southeast Asian Mobile Network Operator
AI-powered multilingual voice and SMS platform scaled daily customer reactivation outreach 10x and more than doubled conversion rates for a Southeast Asian MNO targeting aggressive subscriber growth.
Built on Amazon Bedrock and Amazon EKS
10x
increase in daily outreach capacity
2.1x
improvement in conversion rates
350+
churn micro-segments established
Overview
About the Customer
A rapidly growing Southeast Asian mobile network operator — recently transitioned from MVNO to full MNO with 5G services — serving approximately 1.5 million subscribers in a highly competitive prepaid market. Its strategic focus: accelerate subscriber growth, reduce churn, and build a scalable digital-first engagement model.
The Challenge
Business Challenges
- The operator's goal was to double its customer base within two years. The critical lever was reducing churn and reactivating a large dormant subscriber base that generated limited revenue and showed low responsiveness to traditional campaigns.
- Outreach was manual and telecalling-dependent — capping daily contacts at approximately 1,400 with a 1.4% conversion rate. Customer Value Management capabilities were nascent, with no automated lifecycle engagement and no structured segmentation.
- Earlier AI voice trials exposed gaps in localized language tuning, conversational quality, and operational scalability. Without an AI-driven engagement platform, the operator risked sustained churn and an inability to hit its growth targets.
The Solution
Solution Approach
OBZ designed and delivered an AI-powered multilingual voice bot with SMS integration (English and Bahasa Malaysia) for customer winback, alongside a Foundational CVM Framework to arrest churn — all on Amazon Web Services.
Multi-Channel Communication Platform
The solution delivers a comprehensive Communication Platform as a Service (CPaaS) approach, combining voice and SMS channels for unified customer engagement:
- Voice Channel: AI-powered outbound calling for personalised, conversational customer interactions with natural language understanding and real-time dialogue generation in Bahasa Malaysia and English.
- SMS Channel: Automated SMS delivery for post-call follow-ups, campaign reminders, offer confirmations, and re-engagement messages, orchestrated through the same platform that manages voice outreach.
- Unified Orchestration: Both voice and SMS campaigns are managed through a single AWS-based platform, enabling consistent messaging, coordinated multi-touch campaigns, and consolidated analytics across all customer touchpoints.
This multi-channel approach maximises customer reach and engagement effectiveness by combining the conversational depth of voice with the cost-efficiency and convenience of SMS messaging.
AI-Powered Voicebot
OBZ implemented a carrier-grade, multilingual voicebot platform hosted on AWS. The solution enabled automated outbound calling to re-engage dormant and at-risk subscribers through personalised, natural conversations in Bahasa Malaysia and English. Key capabilities included:
- Personalised win-back scripts and dynamic content generation
- In-call authentication and validation handled by the AI agent
- Voice persona and tone consistency
- Post-call SMS for reminders and follow-ups
- Real-time analytics and campaign monitoring
The platform was deployed as containerised microservices on Amazon EKS, enabling elastic scaling and high availability.
Foundational CVM Framework
To support preventive churn management, OBZ established a structured CVM foundation including:
- Customer segmentation and churn-risk modelling
- Micro-segmentation logic for targeted campaigns
- Campaign planning and execution workflows
- Performance reporting and insight dashboards
This framework created the foundation for lifecycle-based engagement beyond the MVP phase.
Implementation and Support
OBZ provided end-to-end delivery and support, including:
- Solution architecture and security design
- Platform deployment and configuration
- Model tuning and language optimisation
- Integration with telephony and segmentation systems
- Production monitoring and performance optimisation
This ensured operational readiness and reliability throughout the pilot.
AWS Services Used
Built on AWS
Amazon EKS
container orchestration for voice routing, scheduling, and observability
Amazon Bedrock
foundation model inference for conversational intelligence and dialogue generation
Amazon RDS (PostgreSQL)
transactional storage for call metadata, orchestration state, and audit logs
Amazon S3
campaign input storage and post-call analytics data lake
Amazon ElastiCache
low-latency session and call-context caching
Amazon OpenSearch
vector storage for knowledge grounding and contextual retrieval
AWS IAM, KMS, Secrets Manager
identity management, encryption, and credential governance
Amazon CloudWatch and CloudTrail
monitoring, logging, and auditability
Outcome
Business Outcomes
The AI-powered engagement platform delivered measurable improvements aligned with the operator's churn management and reactivation objectives:
- Increased daily outreach capacity from approximately 1,400 to 14,000 contacts per day (10x improvement)
- Improved conversion rates from 1.4% to approximately 3.0%
- Scaled reliably to support up to 20,000 outbound calls per day in production
- Enabled consistent multilingual engagement in Bahasa Malaysia and English
- Established over 350 micro-segments for churn identification and prioritization
In addition, the MVP validated the feasibility of large-scale, automated voice engagement and provided performance baselines for future expansion. While CVM grid deployment was deferred due to a platform migration, the foundational segmentation and governance framework was successfully established.
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
