Rearchitecting insurance customer experience for the MCP generation

AI is transforming how insurers serve policyholders, detect risk, and operate across increasingly complex technology environments. But one question stands above the rest: Who controls the future of your policyholder and employee experiences, you or your technology vendors?

At Avaya, we believe the answer should be you. That is why we are embracing the open future of AI, building the Avaya Infinity® platform to support Model Context Protocol (MCP).

It is the right move for insurers that need to innovate without abandoning trusted systems, sensitive data, established workflows, or the governance requirements that protect their customers and their business.

Read the press release

What is MCP?

MCP stands for Model Context Protocol, an emerging open standard that enables AI models such as GPT, Claude, or Gemini to securely and reliably interact with external tools, data sources, APIs, and user context in a structured way.

For insurers, this can give AI access to the context needed to understand a policyholder, claim, policy, interaction history, or servicing request across multiple systems. Instead of operating in isolation, AI can work with relevant enterprise context, established workflows, permissions, and operational intent.

The result is AI that can become significantly more useful in high-stakes insurance environments where accuracy, security, continuity, and trust matter.

"This is not a 'wait-and-see' moment. Avaya believes the time to be intentional about building the definitive open orchestration engine for the modern enterprise is now."

David Funck, Chief Technology Officer, Avaya

How does MCP work?

MCP is built on a client → model → server architecture: 

  • The Client gathers context such as policyholder information, employee role, current task, permissions, interaction history, or servicing environment.
  • The Model such as Claude or GPT receives the context and a list of available tools or actions.
  • The Server hosts approved tools, APIs, databases, and enterprise systems and executes authorized actions requested by the model.

All communication follows structured schemas so interactions can be interpretable, traceable, and governed, an essential requirement when AI is interacting with sensitive insurance information and business processes.

MCP Illustration on how it works

Avaya Infinity with MCP: Achieving AI-powered insurance communications intelligence

 

Insurers can move beyond static dashboards to analyze CX data, uncover trends, and generate real-time views. Supervisors, claims leaders, and servicing teams can explore vital insights while maintaining full system security and data governance.

Removing friction from insurance AI with MCP

For too long, bringing AI into insurance customer experience has meant choosing between two difficult options:

  • Lock in to a single provider's closed ecosystem, limiting flexibility as AI models, regulations, customer expectations, and business requirements evolve.
  • Create costly, one-off integrations across policy administration, claims, CRM, billing, knowledge, fraud, workforce, communications, and other systems.

This false choice slows innovation and places additional pressure on already complex insurance technology environments. It can also fragment the experience for policyholders and employees, especially when information is distributed across legacy and modern platforms.

Enter MCP, an open, vendor-agnostic standard designed to help AI models interact with tools, data, and operational logic securely and at scale.

Think of MCP as a universal connection layer for AI. Instead of building a different connection between every model and every insurance system, organizations can create a standardized way for approved AI capabilities to access the enterprise context and tools they need.

As the AI ecosystem evolves, this approach can give insurers greater freedom to innovate while maintaining control over the systems, data, workflows, and customer relationships that make their business unique.

Innovate on your terms with an open, model-agnostic platform
  • Choose the right LLM or AI capability for each insurance use case without being locked into a single model provider.
  • Let AI securely access approved information from CRM, policy administration, claims, billing, knowledge bases, fraud systems, data platforms, and other sources where your enterprise truth lives.
  • Future-proof your ecosystem so new models and tools can be introduced without repeatedly rebuilding the integrations that connect AI to your business.
Deliver contextual experiences that strengthen policyholder relationships
  • Create flexible CX environments informed by policyholder history, policies, claims, interactions, intent, and journey context.
  • Give service representatives, claims professionals, and other employees a more complete view across touchpoints, systems, and time.
  • Help employees access the knowledge and context they need to deliver faster, more personalized, and more consistent experiences across voice and digital channels.
Build with enterprise-grade trust, security, and governance at scale
  • Apply zero-trust principles so AI interactions can pass through governed gateways with defined identities, permissions, and access controls.
  • Use auditability and tool-level controls to help support insurance security, privacy, governance, and regulatory requirements.
  • Keep human judgment in the process for sensitive or high-impact actions, helping insurers accelerate work with AI while maintaining appropriate oversight.

MCP enables AI systems to operate with dynamic awareness of customer, employee, session, and business context.

For Avaya Infinity, this means AI models can deliver experiences informed by relevant information across complex insurance environments. A servicing interaction could draw upon approved policy, claim, billing, CRM, knowledge, and interaction data while respecting the permissions governing each source.

That level of contextual intelligence can be especially valuable for insurers managing thousands of employees, workflows, integrations, products, policies, and customer journeys across multiple lines of business.

Shaping the future of open AI for insurance customer experience

Avaya is not just talking about open AI. We are actively helping shape open standards while bringing native MCP support to Avaya Infinity for secure, model-agnostic, enterprise-ready orchestration.

For insurance organizations, that creates three important advantages:

  • Control: Gain more control over AI models and tools so you can select capabilities based on the requirements of each insurance use case instead of the limitations of a single vendor.
  • Governance: Strengthen interoperability and oversight across customer experience systems, data, workflows, and AI so capabilities can be introduced responsibly.
  • Future-readiness: Build an AI architecture designed for complex, regulated environments where security, resilience, auditability, and long-term trust are essential.
Agentic AI Foundation and Linux Silver Member Logo

Collaboration with Databricks

As part of our commitment to secure, scalable, and open AI, Avaya is collaborating with Databricks to bring enterprise-grade governance and data privacy capabilities to the MCP implementation within the Avaya Infinity platform.

For insurers, this approach can help establish controlled access between AI and the structured and unstructured data distributed across the enterprise.

With fine-grained access controls, audit logging, and governed integration with enterprise data sources, insurers can explore AI-powered customer and employee experiences while maintaining greater visibility into how information is accessed and used.

"Generative AI offers tremendous potential to transform customer experiences, and we are thrilled to collaborate with Avaya to help organizations quickly unify their data, simplify data and AI governance and security, and ultimately deliver AI that understands their business."

Heather Akuiyibo, VP of GTM Integration, Databricks

Databricks logo

From fragmented insurance systems to freedom with MCP

A typical insurer may have dozens or hundreds of tools and data sources, including policy administration systems, claims platforms, CRM, billing, knowledge bases, fraud systems, document repositories, data platforms, and workforce applications. At the same time, multiple AI models, copilots, applications, and agents may need controlled access to those systems.

Before MCP, connecting each system to each AI model could require its own point-to-point integration. As the number of systems and AI capabilities grows, so does the development effort, maintenance burden, complexity, and technical debt. MCP changes that paradigm by introducing a standardized intermediary layer. Instead of solving an increasingly complex N x M integration problem, organizations can move toward a more scalable N + M model.

Each enterprise tool can expose approved capabilities through an MCP server, while AI models and applications connect through MCP clients. Once components can communicate through MCP, insurers gain a more flexible foundation for connecting AI with existing and future systems. That can dramatically reduce redundant integration work while protecting investments already made in the technologies and workflows that support policyholders today. This matters because insurance customer journeys rarely live in one application. A single interaction may touch communications, CRM, policy administration, claims, billing, knowledge, authentication, document management, fraud detection, and analytics.

Avaya Infinity with MCP can help connect those technologies, insights, channels, and workflows so insurers can modernize the experience without requiring the entire enterprise architecture to be replaced first.

The potential ROI of MCP for insurers

With MCP, Avaya Infinity users can unlock deeper business value by giving AI access to more relevant context and approved enterprise capabilities.

For insurers, that can mean faster service, more informed claims interactions, more natural conversations, stronger brand consistency, greater employee productivity, and better continuity throughout the policyholder journey.

MCP creates the foundation for generative and agentic AI that is not simply intelligent, but situationally aware of the customer, employee, policy, claim, business process, and permissions surrounding an interaction.

 Before MCPAfter MCP
IntegrationCustom connectors between individual AI models and insurance systemsA standardized MCP interface can work across models and approved enterprise tools
Deployment timeLong, expensive due to bespoke engineering 50–70% faster, lower costs via reuse (quiq.com, Palma AI, Humanloop, aibase.com
Data accessAI operates with limited visibility into customer and operational contextAI can access approved real-time context across CRM, policy, claims, billing, knowledge, and other systems
Risk managementEmployees manually search multiple systems to validate information and identify potential issuesAI can assist with cross-system checks and surface relevant information earlier in the workflow

The future of insurance AI is open

Insurance organizations have spent decades building the systems, processes, data, expertise, and customer relationships that make their businesses unique. They should not have to surrender control of those investments to participate in the AI era. The world's leading AI technology providers are increasingly embracing open, contextual approaches to AI. Avaya shares that vision.

By bringing MCP to enterprise customer experience, Avaya can help insurers connect AI with the systems they already trust, adopt the models that best serve each use case, preserve greater control over their architecture, and continuously innovate as the AI ecosystem evolves.

The future of insurance customer experience should not be defined by a closed technology stack. It should be defined by what your organization can make possible.

Glossary

Full glossary of terms below.

TermDefinitionWhy It Matters for Insurance
MCP (Model Context Protocol)An open protocol that standardizes how AI models interact with tools, APIs, and memory via structured context objects.Creates a scalable way to connect AI with policy, claims, CRM, billing, knowledge, and other insurance systems.
ContextA bundle of structured information passed to a model—such as user ID, session history, roles, time, or tool availability.Helps AI understand the policyholder, employee, claim, policy, and situation surrounding an interaction.
ModelThe generative AI (e.g., Claude, GPT-4, Gemini) that receives the MCP context and generates responses.Gives insurers flexibility to select different AI models based on the requirements of individual use cases.
ClientThe application or interface (e.g., chat window, agent framework, customer portal) that gathers and sends context to the model.Can include customer portals, employee desktops, virtual assistants, agent applications, and other insurance experiences.
ServerThe logic layer that hosts tools, APIs, functions, and data sources the model can use via MCP.Creates a governed way for AI to interact with insurance systems without giving models unrestricted access to enterprise resources.
ToolsDiscrete functions or APIs (e.g., “search knowledge base,” “check ticket status,” “update CRM”) exposed to the model via MCP.Allows AI to move beyond conversation and assist employees or customers with real insurance tasks.
SchemasThe structured definitions (in JSON or YAML) for how tools and context should be represented and passed to models.Creates consistency between AI models and insurance systems, making integrations easier to understand, govern, and troubleshoot.
MemoryPersistent context retained across sessions (user history, preferences, past actions) provided to the model.Can enable more continuous policyholder experiences so customers do not have to repeatedly explain the same situation.
Ephemeral ContextShort-lived, session-specific information (e.g., current task, temporary variables).Helps AI remain focused on the immediate servicing, claims, or employee workflow without unnecessarily retaining temporary information.
AgentAn AI system that can take actions, invoke tools, and make decisions based on MCP context — often used in autonomous or semi-autonomous workflows.Provides the foundation for controlled agentic AI that can assist with multi-step insurance processes and workflows.
OrchestrationCoordinating multiple steps, tools, and model outputs to achieve a higher-order business goal.Helps insurers connect customer interactions with the many systems and processes required to complete servicing and claims journeys.
ObservabilityThe ability to inspect what context was passed, what decisions the model made, and how tools were used.Supports the transparency, governance, troubleshooting, and oversight required in highly regulated insurance environments.
InteroperabilityThe ability for different AI models and platforms to work with a shared structure of tools and context.Gives insurers greater flexibility to evolve their AI strategies without replacing established systems or becoming dependent on a single AI provider.