Overview
Environment
Azure/Hybrid
Engagement type
Funded Assessment/PoC
Duration
2 weeks
Scope
Enhancing contact center experience by integrating Microsoft Copilot with Facets and Dynamics 365 to reduce agent effort, improve response time, and increase first-call resolution
Client leverages Facets as its core system for storing member health insurance details, integrated with Dynamics 365 to display benefits information to contact center agents. This PoC focuses on deploying Microsoft Copilot to act as an intelligent intermediary, dramatically reducing the manual effort agents currently expend navigating multiple systems to locate and relay member benefit information.
Customer snapshot
Industry
Healthcare
Headquarters
Baton Rouge
The business challenge
- Member health insurance details are stored in Facets, a healthcare claims management application, and accessed through Dynamics 365 to view benefits information.
- Agents must navigate multiple dropdown menus in Dynamics 365 to locate the relevant member information.
- Agents must read and comprehend the retrieved information to provide the correct details.
Impact of the problem
- Slower response time and increased effort when retrieving and understanding member information.
- This process slows response time and increases effort when retrieving and communicating member information.
PoC objective
To streamline the process of retrieving member plan and benefit information by enabling agents to use Copilot to fetch member data from Facets, reducing manual navigation and effort.
The Sonata solution – what we did
Sonata designed and implemented a Microsoft Copilot-based intelligent assistant, embedded directly within the Dynamics 365 Customer Engagement (CE) interface. The solution integrates Copilot with the client’s Facets system via API to deliver a seamless, conversational benefit inquiry experience for contact center agents.
Solution highlights
01 Copilot Launch from D365 CE
The Copilot bot is launched directly from the Dynamics 365 CE user interface, keeping agents within a single workspace.
02 Automatic Contextualization
Upon launch, the bot instantly accesses the relevant Member Profile and Product (Plan) Profile from D365, eliminating the need for manual data entry.
03 Dynamic Indexing
A temporary, session-based table is created to index plan summary data, enabling rapid and accurate query responses during the call.
04 Robust Fallback Mechanism
A pre-existing Plan Summary query serves as a reliable backup to ensure data availability even if the primary session table is unavailable
Microsoft technologies used
Microsoft Copilot Studio – Conversational bot design and orchestration
Dynamics 365 Customer Engagement (CE) – Agent desktop and context source
Azure API Management / Connector – Facets API integration layer
Power Platform – Workflow and session-based data management
Microsoft + Sonata credibility
30+ years
of partnership
Microsoft AI Business Solutions Inner Circle member
Microsoft Frontier Firm Partner
How we did it
The Copilot conversation sequence
The solution follows a structured conversation flow that guides the agent through the benefit inquiry process:
| Launch Copilot from D365 CE | Member Profile and Eligibility context are automatically passed to the bot |
| Retrieve eligibility / plan information | Copilot calls Facets API to fetch the member's active plans |
| Agent selects plan | Plans are listed by PDPD ID / Plan Description for agent selection |
| Retrieve product types | Copilot fetches associated product types for the selected plan |
| Agent selects product type | Agent narrows down to the specific benefit area of interest |
| Retrieve benefit summary | Copilot returns the benefit summary data from Facets |
| Dynamic indexing | A session-based table is created and indexed for rapid natural language querying |
| Natural language Q&A | Agent asks benefit-related questions in plain English |
| Copilot returns responses | Accurate, contextual answers are surfaced instantly |
| Agent reviews and relays | Agent confirms and communicates the information to the member |
Results and benefits
The following metrics represent the anticipated impact based on the PoC design and preliminary outcomes. Quantified baselines will be confirmed upon full deployment.
| Metric | Improvement |
|---|---|
| Average handle time | Copilot retrieves all data in one query thereby reducing AHT by up to ~ 30-40% and faster member service |
| CSR Navigation Effort | Conversational interface replaces manual lookup thereby eliminating 100% of manual lookup steps |
| Accuracy of benefits communication | Copilot contextualizes and surfaces correct data leading to reduction in miscommunication errors by up to ~20–25%. |
