ReleaseBytes
aws AWS Contact Center Blog ·

Amazon Connect's Case Management Approach Explained

bloginfraawsengineer
announcement

This post details Amazon Connect's unique approach to case management, differentiating it from traditional Customer Service platforms. It explains how separating the 'container' (Case) from the 'unit of work' (Contacts/Tasks) enables more flexible routing and better context preservation. This architectural decision is presented as crucial for optimizing both current contact center operations and future AI-driven workflows, benefiting engineers and architects managing customer service platforms.

  • →Amazon Connect Separates Work Units from Case Container
  • →Case as Orchestration Layer, Not a Queue Ticket
  • →Parallel and Sequential Routing for Complex Workflows
  • →Improved Context Preservation for Agentic Workflows
  • →Mitigating Stranded Workflows
Enhancements (2) ›
  • Parallel and Sequential Routing for Complex Workflows

    The separation allows for parallel routing of multiple work items from a single case to different specialists, and sequential routing when dependencies exist, all while maintaining case continuity and visibility for the case owner.

  • Improved Context Preservation for Agentic Workflows

    This model is highlighted as beneficial for AI agents, providing them with a structured 'mission brief' that includes customer history, attempted actions, and the ultimate goal, enabling more informed decision-making.

Notes (4) ›
  • Amazon Connect Separates Work Units from Case Container

    Amazon Connect distinguishes between routable work items (contacts and tasks) and persistent case containers. This architectural choice aims to provide a more powerful foundation for evolving contact centers, especially with the integration of AI.

  • Case as Orchestration Layer, Not a Queue Ticket

    Unlike traditional systems where cases are routable objects that move through queues, Amazon Connect treats cases as persistent containers that orchestrate, track, and provide context for multiple interactions related to a customer goal.

  • Mitigating Stranded Workflows

    The architecture prevents 'stranded workflows' by allowing individual work items to be re-routed if an agent is unavailable, without losing case records or SLA tracking, ensuring seamless customer experience.

  • Optimizing Multi-Specialist Collaboration

    For scenarios requiring multiple teams (e.g., billing and fraud analysis), Amazon Connect enables simultaneous work on discrete tasks under a single case, accelerating resolution times compared to sequential processing.

Read the original announcement →

https://aws.amazon.com/blogs/contact-center/rethinking-the-case-how-amazon-connects-approach-unlocks-the-future-of-customer-service/

Related releases