Key results
by automating premium tracking, claims performance, and agent reporting
reducing average processing time from 10 to 8.5 days
in broker-level policy conversions through performance visibility
in customer renewal rates by flagging high-risk churn cohorts
About the Client
A multi-line direct insurance company serving personal and commercial policyholders across United States of America. The client offers auto, health, life, and property insurance, and was seeking to streamline underwriting insights, improve claims management, and strengthen agent-level accountability.
Key Challenges
Disconnected systems for policies, claims, customer service, and agents
Delayed access to critical metrics like premium growth, loss ratios, and churn
Manual reporting slowed down performance visibility and proactive decision-making
Our Approach
- Built a cloud-native data warehouse using Snowflake on AWS
- Integrated data from policy, claims, and CRM systems into unified data marts
- Delivered department-specific Tableau dashboards with predictive layers for churn and fraud
Implemented Solution
Data Engineering & Infrastructure on Snowflake
- Connected policy admin, claims, and CRM systems to Snowflake using CDC and batch APIs
- Designed modular data marts for Policy, Claims, Customer, and Agent with built-in SCD handling
- Implemented monitoring and validation pipelines for real-time data reliability
Insurance-Specific BI Dashboards using Tableau
- Policy & Underwriting Analytics
- Monitored premium written vs. earned across LOBs, regions, and risk segments
- Tracked quote-to-bind ratios and underwriter efficiency
- Identified underperforming policies based on profitability and loss ratios
- Monitored premium written vs. earned across LOBs, regions, and risk segments
- Claims Management
- Visualized claim volume trends, resolution cycles, and escalation patterns
- Tracked claim aging, adjustment timelines, and fraud-prone cases
- Delivered insights into claim leakage and reserve accuracy
- Visualized claim volume trends, resolution cycles, and escalation patterns
- Customer Retention & Churn Analysis
- Highlighted churn-prone cohorts using tenure, claim experience, and payment behavior
- Visualized renewal success by product, geography, and agent
- Integrated NPS, support interactions, and lapse triggers for retention scoring
- Highlighted churn-prone cohorts using tenure, claim experience, and payment behavior
- Agent & Broker Performance
- Compared agent productivity using premium, persistency, and claims ratios
- Enabled leaderboard reports by region, tenure, and product specialization
- Tracked quote follow-ups, policy drop-offs, and renewal effectiveness
- Compared agent productivity using premium, persistency, and claims ratios
Predictive Modelling
- Built churn prediction models using engagement, claims history, and tenure
- Applied unsupervised clustering for agent segmentation and personalized coaching
- Deployed fraud anomaly detection models for early claim intervention
Technologies Used
