activeAviation & Enterprise

Journey - Contact Centre Modernization Platform

A cloud-native contact centre platform delivering 30% cost savings and improved customer experience for a major aviation company through AI-powered automation and intelligent routing.

Impact: 30% cost reduction, 25% improvement in first-call resolution

Role
Principal Solution Architect
Year
2020-Present
Amazon ConnectAWS LambdaAmazon LexAmazon EventBridgeAWS CDKTypeScriptPythonDynamoDB

The Challenge

The legacy contact centre infrastructure was struggling to meet modern customer expectations. Key challenges included:

  • Scalability issues during peak periods like holidays and irregular operations
  • Siloed systems requiring agents to toggle between multiple applications
  • Limited AI capabilities for automated customer service
  • High operational costs with aging on-premises infrastructure

The Solution

We designed and implemented a cloud-native contact centre platform built on Amazon Connect, with AI capabilities powered by Amazon Lex and a serverless backend architecture.

Key Architectural Components

Intelligent IVR: Natural language understanding enables customers to describe their needs in their own words, routing to the right resource without frustrating menu trees.

Agent Desktop Integration: A unified interface pulling data from booking, loyalty, and customer history systems—reducing average handle time and improving agent satisfaction.

Real-time Analytics: Streaming data to a data lake enables operational dashboards and feeds machine learning models for continuous improvement.

Automated Rebooking: During irregular operations, customers can complete rebooking through automated channels, reducing queue times for those who need human assistance.

Results

  • 30% reduction in cost per contact
  • 25% improvement in first-call resolution
  • 40% of routine inquiries handled through automation
  • Scalable infrastructure that handled 3x normal volume during major disruptions

Technical Decisions

The project involved numerous architectural decisions documented in ADRs. Key choices included:

  • Serverless compute (Lambda) for business logic to eliminate scaling concerns
  • Event-driven architecture using EventBridge for system integration
  • Infrastructure as Code with AWS CDK for reproducible environments
  • Multi-region deployment for disaster recovery and latency optimization

Lessons Learned

  1. Start with a narrow scope and expand incrementally
  2. Invest heavily in data quality and integration
  3. Agent adoption is as important as technical architecture
  4. Build observability in from day one

Related Projects

completed

Serverless Migration - Legacy Systems to AWS

Led the migration of critical aviation operations systems from on-premises infrastructure to a serverless architecture on AWS, achieving 40% performance improvement.

Impact: 40% performance improvement, 50% infrastructure cost reduction

Role: Lead Architect

AWS LambdaAmazon API GatewayAmazon Aurora ServerlessAWS Step Functions+2 more

Building something similar?

I enjoy comparing notes on products, platforms and the systems behind them.

Get in Touch