NewNew: The enterprise guide to Agentic AI — 24 min read.

Read →
Case study · Health Payers · Agentforce

46% member-services deflection for a Medicare Advantage payer — CMS-safe agentic AI

A regional Medicare Advantage payer was buckling under AEP call volume for benefits, PCP changes, ID cards and prior-auth status. pronix.ai deployed Salesforce Agentforce on Health Cloud with Amazon Connect voice and CMS-safe guardrails — 46% deflection and CSAT flat through AEP.

Client
Regional Medicare Advantage payer, 1.2M members
Industry
Health Payers
Platform
Salesforce Agentforce · Salesforce Health Cloud · Amazon Connect
By pronix.ai CX Engineering8 min readQ1 2026
46%
Deflection on top-5 intents
-11pt
Abandon rate through AEP
+0.2pt
Member CSAT
0
CMS marketing findings

*Representative outcome; results vary by client, scope and platform configuration.

The challenge

AEP week-one call volume was 4.2x steady state and abandon rates were pushing 18%. Any assistant had to respect CMS marketing rules, never quote plan pricing outside the sanctioned language, and route anything clinical to a licensed agent.

Our approach

Step 01

CMS-safe intent boundary

Locked the assistant to five non-marketing intents — benefits summary, PCP change, ID card, prior-auth status, claim status. Anything else routed to a licensed agent within one turn.

Step 02

Health Cloud grounded answers

Agentforce topics grounded on member record, plan document library and prior-auth queue — no free-text drift, every answer traceable to a sanctioned source.

Step 03

Voice + web from day one

Same Agentforce topics served the Amazon Connect voice bot and the member portal chat; one authoring surface, two channels live together.

Step 04

AEP surge playbook

Pre-loaded 12 seasonal FAQs, added an overflow callback offer at 60-second wait, and staged a war room for the first ten AEP days.

AEP week one used to be the worst week of our year. This year the queue held — and members got answers at 10pm.

VP Member Services
For VP Member ServicesFor Chief AI OfficerFor CIO

Illustrative case study. Scenarios, metrics, quotes and client details are representative composites based on Pronix engagements and industry benchmarks unless a named client is shown with written consent. Outcomes vary by client, scope, data quality and platform configuration. Nothing on this page is a guarantee, warranty or professional advice. See our Terms of Use for the full disclaimer.

Trademarks.

AWS, Salesforce, Microsoft, NICE, Genesys, Kore.ai, Google, ServiceNow, Amazon Connect and logos referenced on this site are trademarks of their respective owners. References are for descriptive purposes only and do not imply endorsement, sponsorship or partnership beyond stated partner relationships. See our Disclosures.

Bring this to your program

Talk to the team that shipped it.

Book a working session with a pronix.ai lead on this practice — we'll map the approach to your platform, industry and constraints.