New Lyzr launches Control Plane for AI Agents Access now (opens in a new tab)
WTW Automated B2C Billing Operations with Lyzr
Case Study · B2C Billing with Lyzr

WTW Automated B2C Billing Operations with Lyzr

Agentic billing automation Billers stay in control A pipeline that learns over time
Try Agent Talk to Us
About the company

Willis Towers Watson (WTW) is a leading global advisory, broking, and solutions company. Their billing operators process thousands of activities daily inside EPIC, their broker-management system of record — each requiring manual data pulls from five separate systems, hand reconciliation, and a billing transaction post.

The challenge

The problem statement

01

Five systems, zero automation

Billers manually pulled and reconciled data across EPIC, RMS, Industry Mapper, IRD, and CABR — with constant context-switching and no unified view.

02

High manual effort, every activity

Reading attachments, hunting for policies, and cross-checking amounts consumed significant time — repeated thousands of times over.

03

Inconsistent quality across billers

Every biller worked slightly differently. Errors surfaced weeks downstream, expensive to correct and hard to trace.

04

No audit trail

No clean, field-level record of who changed what and when — a compliance risk on every activity.

The approach

How Lyzr solved it

Lyzr implemented an agent-driven inbound SDR system that combined structured knowledge, automated decisioning, and human-in-the-loop controls to manage inbound engagement at scale.

Lyzr built an agentic billing pipeline

~22 AI agents fan out across all five systems, extract and validate every data point, and present the biller with a single confidence-scored recommendation. Nothing posts to EPIC without human approval.

Lyzr unified five systems into one workflow

One interface pulls from EPIC, RMS, Industry Mapper, IRD, and CABR — reconciled automatically, no context-switching required.

Lyzr gave billers control

Billers can correct any step and re-run only what's downstream. Every edit is logged with user, timestamp, and notes.

Lyzr built a full audit trail

Field-level edit history captured at every step — exportable and available for compliance review on demand.

Results

The outcome

Billers became decision-reviewers

Extraction and reconciliation happen automatically. Billers review, approve, or correct — in a fraction of the previous time. Target: 70–80% of activities processed straight through.

Errors caught before posting

Automated check groups flag exceptions before anything reaches EPIC — replacing the variance that came from each biller working differently.

The system improves with use

Corrections feed back in. Auto-rates rise, manual corrections fall, and the pipeline gets more accurate with every activity.

Compliance gets a clean audit trail

Every action, correction, and approval is logged — giving finance and compliance a complete record for any billing decision.

How Lyzr builds enterprise AI on AWS
Behind the scenes

How Lyzr builds enterprise AI on AWS
Security

How Lyzr handled security

01

Enterprise-grade compliance

Azure AD / JWT authentication, CORS allow-list, rate limiting, and managed-identity database access built to enterprise standards from day one.

02

Compliance-aware logging

Full prompts are not logged in production. The audit trail captures diffs and user actions without exposing sensitive billing data.

03

Controlled rollout by design

Posting to EPIC is feature-flag gated running in recommendation-only mode until the team is ready to enable automated posting.

More stories

Similar case studies