How Toyota Insurance cut customer service costs by 98.5% and achieved 60% self-service with Amazon Connect

Toyota Insurance Management Solutions (Toyota Insurance) is a technology-driven insurance agency serving predominantly Toyota vehicle owners across North America. As a forward-thinking organization, Toyota Insurance continuously seeks innovative ways to enhance customer experience while maintaining operational efficiency. Back in February 2025, Toyota Insurance faced a critical decision: their SaaS-based web chat solution for Toyota Auto Insurance, their flagship product, was failing to meet customer expectations. Policyholders were experiencing wait times for even simple questions, and the system couldn’t provide assistance outside business hours. The legacy solution, which provided support for customer-facing insurance services including policy updates, billing inquiries, and general inquiries, was no longer serving customers effectively. With roughly 3,000 chats per month, the company needed a solution that would give Toyota Insurance customers instant, 24/7 access to accurate information.

In this post, we explore how Toyota Insurance successfully migrated from a legacy SaaS-based web chat platform to Amazon Connect in just one month.

This transition transformed the customer experience with 24/7 instant self-service, now handling 60% of inquiries, while also reducing costs by 98.5% and laying the foundation for continuous improvement through AI-driven analytics.

The challenge: Limitations of a SaaS-based solution

Toyota Insurance had been using a third-party SaaS solution for web chat support. While it initially served the basic need to connect customers with support agents, several limitations became increasingly apparent:

1. Poor customer experience during peak times: With nearly 0% self-service deflection, every inquiry required an agent, creating wait times that frustrated Toyota Insurance customers seeking quick answers to simple questions
2. No after-hours support: The system couldn’t effectively serve customers outside business hours, forcing policyholders to wait until the next business day for routine inquiries
3. High operational costs: The monthly license-based pricing model made the solution prohibitively expensive as chat volumes grew, with costs scaling regardless of actual usage
4. Inability to improve systematically: Analyzing conversation transcripts and identifying improvement opportunities was cumbersome, making it difficult to learn from customer interactions and proactively address common concerns

As Toyota Insurance leadership evaluated alternatives, their primary goal was clear: deliver the exceptional customer experience that Toyota Insurance policyholders deserve. They sought a solution that would provide instant, accurate answers 24/7 while enabling continuous improvement through the kaizen methodology, the philosophy of continuous improvement that is central to Toyota’s culture.

The solution: Amazon Connect with generative AI

After engaging with the AWS account team, Toyota Insurance discovered that Amazon Connect offered exactly what they needed: an AI-native omnichannel cloud contact center with deep integration across the AWS ecosystem. The migration plan included:

Core Amazon Connect Implementation

• Amazon Connect contact center: Cloud-native customer experience solution providing omnichannel capabilities
• Amazon Q in Connect: Generative AI-powered agent assists with retrieval-augmented generation (RAG) for knowledge base articles, offering the flexibility to choose foundation models, customize system prompts, and configure guardrails for safe and accurate responses
• Self-service chatbot: Intelligent chatbot leveraging Amazon Q in Connect to handle common customer inquiries without agent intervention

Data analytics and Kaizen pipeline

Building on Amazon Connect’s native integration with AWS services, Toyota Insurance implemented a sophisticated analytics pipeline:

1. Amazon Kinesis Data Streams: Captures Contact Trace Records (CTR) in real-time as conversations occur
2. Amazon S3: Stores CTRs and full conversation transcripts for analysis
3. AWS Glue and Amazon Athena: Creates queryable tables from the raw data stored in S3
4. Amazon SageMaker AI Studio: Applies custom domain-specific machine learning models including:
   • Embedding models to understand semantic content
   • Clustering algorithms to identify common themes and issues
   • Natural language processing for insight extraction
5. Kaizen implementation: Automatically generates improvements based on analysis:
   • New knowledge base articles for common customer questions
   • Enhanced system prompts to improve response accuracy and tone
6. Continuous feedback loop: Updated Knowledge Base (KB) articles and refined system prompts are deployed back to Amazon Q in Connect

This architecture enables Toyota Insurance to practice true kaizen, continuous improvement driven by data and insights.

Figure 1: Toyota Insurance Knowledge Base Kaizen Workflow Architecture

The migration: One month from start to launch

One of the most impressive aspects of Toyota Insurance’s journey was the speed of migration. Working closely with the AWS account team, Toyota Insurance completed the migration in just one month. Key success factors included:

• AWS training and workshops: The AWS team provided comprehensive training sessions covering Amazon Connect, Q in Connect, and integration patterns
• Proactive support: AWS Solutions Architects and Technical Account Managers were readily available to address questions and provide architectural guidance
• Clear documentation: AWS’s extensive documentation and best practices accelerated development

“The level of support we received from AWS was exceptional,” notes Yuibi Fujimoto, Chief Data Scientist at Toyota Insurance. “They didn’t just provide a platform; they partnered with us to ensure success.”

The results: Dramatic cost savings and enhanced customer experience

The results of Toyota Insurance’s migration exceeded expectations across multiple dimensions:

98.5% cost reduction

By moving from the SaaS-based solution to Amazon Connect, Toyota Insurance achieved a 98.5% reduction in customer service infrastructure costs. These dramatic savings came from several factors:

• Pay-as-you-go pricing: Amazon Connect’s consumption-based model eliminated the fixed monthly license fees, allowing Toyota Insurance to pay only for actual usage
• Reduced agent time: The effective self-service deflection meant fewer conversations required human agents, further lowering operational costs
• Elimination of vendor markups: Direct access to AWS services removed the markup typically associated with third-party SaaS solutions

60% customer inquiry deflection rate

Perhaps the most impressive metric is the evolution of Toyota Insurance’s deflection rate:

• Before migration: Nearly 0% – virtually every inquiry required agent intervention
• Week of launch: 34% – the self-service chatbot immediately began handling one-third of inquiries
• Current: 60% – through continuous improvement and KB article and system prompt refinement

This transformation has fundamentally improved the customer experience for Toyota Insurance policyholders:

• Immediate assistance, anytime: Customers no longer wait in queue for common questions about policy details, billing, or coverage. The intelligent self-service chatbot provides instant, accurate answers 24/7. This is particularly valuable for busy Toyota Insurance customers who need quick answers outside traditional business hours.
• Better agent interactions: For the 40% of inquiries that do require human assistance, customers benefit from shorter wait times and more focused attention. Agents now handle genuinely complex situations, such as claims guidance, policy customization, or unique coverage questions, where their expertise adds the most value.
• Consistent, accurate information: The RAG-based approach ensures every customer receives the same high-quality, up-to-date information, reducing confusion and follow-up inquiries. Toyota Insurance policyholders can trust they’re getting accurate answers whether they interact with the chatbot or an agent.

Easy customization and maintenance

Unlike the previous SaaS solution, Amazon Connect’s RAG-based approach makes it simple to:

• Add new knowledge base articles as new scenarios emerge
• Update existing content based on changing policies or regulations
• Refine system prompts to improve response quality, accuracy, and brand voice
• Customize responses for insurance-specific terminology and compliance requirements
• Continuously optimize both content and conversation behavior based on real customer interactions

Looking forward: Scaling AI-driven customer service

With the success of their Amazon Connect implementation, Toyota Insurance is exploring additional capabilities:

• Beyond service chats: Making the chat available for customer acquisition, such as during quotes, potentially in more channels
• More products: Expanding the solution to more than just their flagship product, Toyota Auto Insurance

“This migration wasn’t just about cost savings. It was about building a foundation for the future,” notes James Eaker, Director of Contact Center Sales and Customer Care at Toyota Insurance. “We now have a platform that will continue to deliver value as our needs evolve and AI capabilities advance.”

Conclusion

Toyota Insurance’s journey from a legacy SaaS-based chat solution to Amazon Connect demonstrates the transformative potential of cloud-based, AI-native customer service infrastructure. Key takeaways include:

• Speed to value: With proper AWS support, complex migrations can be completed in weeks, not months
• Cost efficiency: Cloud-based pricing models can deliver dramatic cost savings compared to traditional SaaS solutions
• AI-powered deflection: Generative AI with RAG can achieve deflection rates of 60% or higher
• Continuous improvement: Integrating ML analytics enables true kaizen-style continuous improvement
• Compliance confidence: AWS’s security and compliance capabilities meet stringent insurance industry requirements

For organizations considering a similar migration, Toyota Insurance’s experience shows that the combination of Amazon Connect and AWS’s broader analytics and ML services offers a powerful platform for modern customer service.

If you’d like to learn more about Amazon Connect and how it can transform your customer service operations, visit the Amazon Connect product page or contact your AWS account team to discuss your specific requirements.

Ready to transform your customer service experience with Amazon Connect? Contact us

About the Authors

Yuibi Fujimoto is the Chief Data Scientist at Toyota Insurance Management Solutions, where he develops innovative data/AI-driven solutions to address complex business challenges. Outside of work, he enjoys spending quality time with his family, exploring new cuisines through cooking, and traveling to learn new cultures.

Genta Watanabe is a Senior Technical Account Manager at AWS. He spends his time working with strategic automotive customers to help them achieve operational excellence. His areas of interest are machine learning and artificial intelligence. In his spare time, Genta enjoys spending quality time with his family and traveling.

Mike Cairns is a Senior Solutions Architect for AWS, specializing in Customer Experience. With over 20 years of experience, he designs and implements solutions for complex business challenges across diverse industries. Leveraging his technical expertise in AWS, he helps organizations modernize infrastructure, improve operational efficiency, and unlock new capabilities.