Capital Rx’s prescription to scale Snowflake with Slingshot

Presented at Snowflake Summit 2025 by Paul Furman, VP Data Management, Capital Rx and Ted Beezy, VP IT, Capital Rx

As an enterprise health technology company and health benefits administrator, Capital Rx leverages Snowflake as the centralized data foundation for Judi®, its proprietary enterprise health platform, to deliver data insights at scale across clinical and analytical teams. 

The company’s data demands soared and Judi’s feature set expanded, so Capital Rx turned to Capital One Slingshot to optimize resources, reduce compute costs and unify warehouse management. Let’s take a look at how Slingshot helped optimize data management for Capital Rx, leading to better patient outcomes through advanced analytics.

With the goal of ensuring affordability and transparency in healthcare, Capital Rx makes providing high-quality healthcare data and insights central to the organization’s purpose. Capital Rx was founded in 2017 as a Pharmacy Benefits Manager (PBM), a middleman that negotiates with pharmaceutical companies and retail pharmacies and processes prescription drug claims on behalf of payers. Customers such as health plans or employers, including health systems, municipalities, labor unions and others, rely on Capital Rx to handle administrative tasks like plan design, eligibility and enrollment, card issuance and formulary management (the list of prescription drugs available to members).

Soon, the company realized the need to build its own application for scalable and secure healthcare claim administration. Launched in 2021, Judi is a cloud-based PaaS enterprise health platform that powers the back-end of claims administration. It’s a PaaS software that unifies all PBM-related workflows and has recently evolved to include medical claim administration.  

The organization uses Snowflake as its centralized data warehouse for managing, maintaining and analyzing healthcare data from Judi. Through Snowflake, the company provides reliable, centralized data to both analytical and clinical workers. Snowflake stores the reusable data models and data transformations, enabling teams to retrieve reliable data quickly for reporting, analytics and data-driven decision making. At scale across the organization, Snowflake allows for ELT and ETL workflows, operational reporting, analytics and insights and reporting. Capital Rx gains high-quality healthcare data insights through Snowflake queries, exports, Tableau and Streamlit integrations.

As Capital Rx’s business grew, the number of Snowflake users increased along with Snowflake usage in the form of schemas, tables and queries. The organization faced scalability and efficiency challenges in managing quickly escalating data costs and warehouse sprawl. At the beginning of 2024, the volume of cases was set to rise dramatically and the company projected a 300% year-over-year increase in Snowflake costs. With more than two dozen warehouses and unpredictable compute costs, the organization needed greater visibility and control over its data costs and usage.

Capital Rx initiated cost optimization efforts with the objective of reducing costs, maintaining performance and minimizing user overhead. They aimed to address the following:

• Cost management at scale: Snowflake usage, such as data volumes, queries and compute time, had experienced a four-fold growth in a year with the expectation of adding more workloads to the platform. The existing price per query (PPQ) target was untenable with usage rapidly growing and costs rising.

• Performance: The upstream platform required better performance during peak usage. Although the organization aimed to optimize and reduce Snowflake costs, such as through resizing warehouses or consolidating workloads, these efforts needed to avoid any slowdowns or degradations in the system performance.

• User overhead: Many workloads ported to Snowflake were created by business stakeholders. With these queries running at scale in Snowflake, they needed to be operationalized, which created additional overhead for data engineering teams.

The first steps the company took were to build monitoring dashboards that visualized Snowflake usage, bringing greater transparency to its teams. The organization also used Snowflake’s historical load (e.g., from WAREHOUSE_LOAD_HISTORY) to build simple time-based schedules, enabling a  right-sizing of warehouses. Capital Rx also conducted manual reviews of queries while also reaching out to data stakeholders and refining data needs. Using QUERY_TAG or warehouse tags also allowed the company to attribute costs to teams and projects.

Although these internal steps led to improvements, such as reductions in total credits and costs per user, much of the effort was manual, i.e., it was inefficient and time-consuming for data teams. 

Capital Rx sought out a more scalable, automated and efficient way to handle growing data and user demands. The teams needed visibility and control over their Snowflake usage to build predictability into their costs. The organization turned to Capital One Slingshot, a data management tool that provided Capital Rx with new insights into warehouse usage patterns and recommendations for optimization.

With a rapidly expanding Snowflake environment, Capital Rx adopted Slingshot to optimize credit consumption in the following ways:

• Consolidated warehouses: With Slingshot’s recommendations and visualizations, Capital Rx was able to perform a detailed analysis of usage patterns and reduce the number of active warehouses. These efforts included consolidating underutilized warehouses.  To consolidate warehouses according to usage patterns, Capital Rx categorized workloads into three usage types: ETL, Analytics and Transactional, each with distinct performance and cost requirements. Warehouse sizes were matched to the workload patterns (e.g., large clusters for ETL jobs, smaller clusters for transactional queries). 

• Cut costs without sacrificing performance: Capital Rx successfully reduced monthly warehouse spend without any impact on performance. For example, efficiency gains came from ensuring each warehouse configuration is in line with changing workloads. Slingshot adjusts the warehouse size to the customer’s workload requirements. 

• Automated scheduling: Using Slingshot’s dynamic scheduling capabilities, Capital Rx  was able to reduce costs by aligning Snowflake warehouse usage with historical demand patterns. Slingshot analyzed past usage to recommend when to move from a smaller warehouse to a bigger warehouse or vice versa and the optimal auto-suspend time.  

In addition to Slingshot and Snowflake, Capital Rx uses Apache Airflow and dbt in its data environment to automate the orchestration, transformation and delivery of critical reports and insights. Used together, these technologies drive data-based decision-making across key departments. Lastly, deploying Streamlit applications has allowed Capital Rx to distribute interactive, real-time operational reports across the enterprise, broadening data access.

After Slingshot went live in July 2024, Capital Rx immediately experienced cost optimization results. The teams saw a 59% decrease in average monthly Snowflake costs  from July to December, compared to the first quarter of the year (January to March). A detailed analysis to rationalize how warehouses performed in conjunction with the Slingshot implementation led to a 40% reduction in the total number of active warehouses. 

While total credits consumed fell, the number of queries and active users both saw marked increases after the launch of Slingshot. At the same time, the cost per user and cost per query both declined rapidly. 

Using Slingshot, Capital Rx was able to minimize unnecessary compute usage without hurting performance. The health benefits administrator realized its goals to optimize resources while reducing compute costs and unifying warehouse management. 

Today, by leveraging Snowflake as its foundational data platform and Slingshot as a powerful tool, Capital Rx is able to produce advanced analytics for improvements in medical and pharmacy claim adjudication, prescription drug pricing and patient care.