Stephens Insurance, a Little Rock-based insurance brokerage, has launched a proprietary claims analytics platform designed to enhance how clients manage property and casualty risk. The tool, known as PCAT (P&C Claims Analytics Tool), began as an in-house project in 2022, developed by Stephens' risk managers and loss control team. In 2024, it was scaled into a fully operational system now used by most large, complex accounts with significant self-insured exposure.
The pitch is straightforward: most brokers offer static, one-size-fits-all dashboards. PCAT, by contrast, tailors its analysis to the client’s specific business model, whether a grocery chain tracking claims per store, a hospitality group tracking claims per hotel, or a trucking fleet tracking claims per division. Using Microsoft PowerBI, clients and consultants can slice claims data by location, unit, or claim type, run real-time simulations of potential losses, and model different deductible or captive structures.
“Clients needed an avenue to connect with their own data in ways that reflect how their businesses actually operate,” said Lizzie Sanford, Chief Strategy Officer, who guided the team’s transition from an Excel prototype to a scalable, PowerBI-based platform. “It’s not just about visualizing data; it’s about using the data to make strategic decisions — in the moment, with confidence.”
This interactive approach resets the pace of insurance decisions. No more waiting for loss runs at renewal; risk managers and CFOs can explore “what-if” scenarios with real-time claims data year-round. In minutes, teams can stratify historical losses, apply loss-development factors and calculate the total cost of risk, turning analysis into action that improves leverage with underwriters.
Built for Use, Not Just for Show
The platform was designed to move beyond the “check-the-box” approach to analytics often seen in the insurance industry. Every dashboard and metric included in the tool is actively used, either by Stephens internally to manage risk or in direct response to client requests. “We’re not creating charts just for the sake of creating charts,” said J.R. Bizzell, a Property & Casualty advisor and key contributor to the platform’s design and content. “If the dashboard is there, it’s because someone needed it to make a decision — and others probably will too.”
The timing is apt. Property and casualty premiums are rising, reinsurance markets have become more volatile, and coverage layers often have gaps. Some blame climate risk, while others point to regulatory uncertainty and shifting interest rates. Whatever the cause, insurers are demanding that premiums reflect a more uncertain risk outlook. PCAT’s output helps position clients in this market. By grounding renewal conversations in credible loss data and benchmarking, consultants can quantify the premium levels underwriters should target and challenge pricing that doesn’t reflect the actual risk. This has proven especially important for clients evaluating increases in retention levels.
“You’re helping executive teams feel more confident about their decisions,” Bizzell said. “When they see their own data, clearly presented, and it shows that certain risks haven’t materialized for years — that changes the conversation.”
Harnessing Advances in Technology
Behind the scenes, Stephens Insurance continues to invest in infrastructure that can efficiently process claims data. At the core is a custom-trained large language model (LLM) that reads the narrative descriptions of claims and classifies them into loss-control categories, such as slip/trip/fall, repetitive motion, and so on. The model achieves high-90s accuracy and processes data at speeds an order of magnitude faster than the manual process it replaced. The accuracy and consistency of this categorization model enhances Stephens Insurance’s ability to understand key drivers of claims. Additionally, Stephens hosts the model on-premises, reflecting a strong commitment to data governance.
Looking Ahead
The Stephens team envisions future iterations of PCAT that will help predict emerging risks before they materialize. But the foundation — built on relevance, simplicity, and adaptability — is already in place. By continuing to refine the platform based on direct client input, the company aims to provide a more precise and proactive approach to risk management. “We created this platform to power our Property & Casualty risk management strategy,” said Sanford. “It’s what we use to make decisions. And it’s built for the people who need that data to be useful — not just available.”
About the Experts
