AWS re:Invent 2025 - Cochlear enhances contact center observability with Amazon Connect (BIZ215)

Enhancing Contact Center Observability with Amazon Connect

Cochlear's Journey

• Cochlear is a world leader in implantable hearing solutions, providing lifetime hearing outcomes for customers
• Cochlear has a global footprint, operating in over 180 countries and having implanted over 700,000 people
• Cochlear heavily invests in R&D, spending 12% of revenue on it, and partners with AWS to continue innovating

Migrating to Amazon Connect

• Cochlear needed a standardized global contact center to replace its disparate systems across 14 countries
• Key requirements included increased reliability, flexibility to handle diverse call flows and languages, and better visibility/observability
• Cochlear used various AWS services alongside Amazon Connect, such as Lambda, Dynamo DB, Polly, and Salesforce integration
• The migration took about 6 months and resulted in 100% stability in the US contact center with no outages during operating hours

Enhancing Observability and Insights

• Cochlear leveraged Amazon Connect's built-in features like call recording, transcription, and sentiment analysis
• Custom vocabulary allowed Cochlear to improve transcription accuracy for brand terminology and product names
• Searchable call transcripts enabled Cochlear to uncover trends and issues across all customer interactions
• Cochlear built custom reports and dashboards using Amazon QuickSight to consolidate data from various sources

Automated Quality Evaluations

• Cochlear moved from manually evaluating 5 calls per agent per month to automating 22,000 evaluations per month
• The automated evaluations use both rule-based categorization and generative AI-powered questions
• Rule-based evaluations are used for compliance checks and policy violations, while generative AI is better suited for soft skills and holistic performance
• The automated evaluations provide detailed insights, reasoning, and scoring to enable targeted agent training

Forecasting, Capacity Planning, and Scheduling

• Cochlear leveraged Amazon Connect's forecasting and scheduling capabilities, including 15-minute interval scheduling for better agent flexibility
• The machine learning-powered forecasting and capacity planning enabled Cochlear to accurately predict call volumes and staffing needs, even during seasonal fluctuations
• Cochlear was able to implement these capabilities with just 6 months of historical data, as recommended by Amazon Connect

Future Innovation

• Cochlear is experimenting with interactive IVR using Lex and its internal AI "Clark" to enable self-service and seamless escalation to agents
• Cochlear is also exploring integrating Salesforce with Amazon Connect to leverage real-time AI agent assistance and automated case updates
• Cochlear emphasizes the importance of testing new AI-powered capabilities internally before rolling them out to customers

Key Takeaways

• Cochlear was able to achieve 100% stability in its US contact center by migrating to Amazon Connect
• Automated quality evaluations enabled Cochlear to scale its quality management program from 5 calls per agent per month to 22,000 per month
• Cochlear leveraged Amazon Connect's forecasting and scheduling capabilities to improve agent productivity and customer wait times
• Cochlear is continuously innovating by experimenting with interactive IVR, Salesforce integration, and other AI-powered capabilities

Technical Details

• AWS services used: Amazon Connect, Lambda, Dynamo DB, Polly, Salesforce integration
• Automated quality evaluations leveraged both rule-based categorization and generative AI-powered questions
• Forecasting and scheduling utilized 15-minute interval scheduling and machine learning-powered predictions
• Cochlear is exploring integrations with Lex and its internal AI "Clark" for interactive IVR

Business Impact

• Cochlear was able to cut its contact center costs in half by migrating to Amazon Connect
• Improved observability and insights enabled Cochlear to target agent training and optimize its contact center operations
• Automated quality evaluations and forecasting/scheduling capabilities helped Cochlear improve agent productivity and customer experience

Examples

• Cochlear used custom vocabulary in Amazon Connect to improve transcription accuracy for brand terminology like "Cochlear" and product names
• Cochlear built custom reports and dashboards in QuickSight to consolidate data from various sources, including post-call surveys
• Cochlear is experimenting with an integration between Lex, its internal AI "Clark," and Amazon Connect to enable self-service and seamless agent escalation for its sales team