As one of the largest healthcare companies in the United States, CVS Health generated $357.8 billion in revenue in 2023, serving more than 100 million people a year through insurance, retail and pharmacy operations. With this vast footprint, the company's internal application of AI is important not only to improve operational efficiency, but also to provide more personalized care and public health services at scale.
CVS Health consistently states its commitment to integrating data science and AI at the core of its business operations. In particular, the company positions AI as a key factor in improving patient outcomes and reducing friction across pharmacies and clinical workflows. In recent years, this includes the deployment of extended intelligence tools to leverage AI for personalising drug support programs and to support national testing and vaccine distribution efforts.
In the next section, we will explore how CVS Health effectively deploys AI in two core business areas.
- Personalizing your drug workflow to improve patient outcomes: Machine learning models analyse patient behavior and prescription data to enable tailored outreach strategies that enhance medication adherence and streamline pharmacy workflows.
- Extended intelligence to streamline testing and vaccine delivery: AI-driven forecasting tools help optimize staffing and resource allocation for public health initiatives and increase responsiveness during national testing and vaccination campaigns.
Personalizing pharmaceutical workflows to improve patient outcomes
CVS Health leverages AI to address long-standing challenges in healthcare: non-adherence with medicines. Patients often do not take their prescriptions as directed, with serious consequences.
Peer-reviewed studies from 2018 have been published Risk Management and Health Care Policy Non-adherence was estimated at approximately $52.84 billion per year, with per patient costs ranging from $5,271 to $52,000 or more per year.
Furthermore, only about 50% of chronically ill patients in developed countries follow the prescribed treatment plan, according to a World Health Organization report. These figures show the magnitude of the business problem and highlight the opportunities for AI-driven personalization to improve both results and cost-effectiveness.
To target this issue, CVS Health implemented a personalization engine using the Databricks Lakehouse platform. The system features a machine learning model that analyzes a wide range of patient data, including:
- Prescription history
- Insurance claim
- Health status
- Demographics
- Real-time pharmacy interactions
Databricks' use case documentation directly argues that models predict the specific needs and behavioral patterns of each patient to determine which types of support are most effective.
The short section of the following video on broader partnerships of Databricks, CVS, and Azure on these systems focuses on customer behavior and experiences throughout the development process.
Rather than providing uniform reminders, AI systems can provide tailored communications ranging from automated text to pharmacist-driven calls where CVS matches patient preferences and predicted responsiveness. for example:
- Some patients may receive refill reminders tailored to typical pickup actions.
- Others can flag for intervention if they consistently delay prescribing chronic diseases.
In this way, CVS aims to match the intensity and type of outreach to the actual risk level and communication preferences of individual patients.
The use cases are well integrated into the core pharmacy workflow, as CVS claims to be actively using this system to promote the omnichannel pharmacy experience on digital, in-store and mobile platforms.
Although CVS has not released detailed results metrics, the company claims that AI-driven personalization tools are helping to improve patient compliance and automate pharmacist outreach, thereby reducing the manual burden on clinical teams.
According to Databricks, the system has improved patient adherence by 1.6%. Although it is seemingly modest, even minor improvements are important, especially given that non-compliance is estimated to cost the US healthcare system $52.84 billion a year. These measurable improvements demonstrate how large-scale, targeted interventions have a significant financial and clinical impact.
The effectiveness of this model is attributed to its ability to unify siloed data sources and focus on short-term behavioral patterns as well as static demographic profiles. Uniform Silo provides care teams with a clearer view of which patients are at risk and how best to support them without wasting time and resources with interventions of all sizes.
Extended intelligence to streamline testing and vaccine delivery
During the Covid-19 pandemic, CVS health was tasked with rapidly scaled testing and vaccination services across thousands of retail and clinical locations in the United States.
To manage the complexity of this nationwide deployment, the company adopted an enhanced intelligence platform built on Datarobot.
A brief video of the data robot explains the various challenges from acquisition of public trusts to oversee leadership communications early in the project.
The main function of the platform was to generate demand forecasts and staffing plans that could be adapted to local conditions. These AI models were drawn from multiple internal and external data sources, including:
- Predicting CDC infection rate
- State and federal testing obligations
- Local appointment trends
- Workforce availability
- Inventory and Supply Chain Signals
By aggregating these inputs, the AI system has developed recommendations for optimal site stuffing, kit distribution testing, and vaccine inventory planning. All of these were dynamically updated when new data became available.
These customer recommendations allowed resumes as follows:
- It predicts fluctuations in demand at each location up to that day.
- Assign a mobile test unit or additional staff to the need area.
- Reduce latency and use critical materials to avoid excess or undersupplying sites.
Vaccine cold chain logistics overview – from storage to management. (Source: Paho/Who)
Unlike standard predictive models that often rely on strict historical baselines, CVS' AI approach allowed for proactive adjustments. The system's enhanced intelligence design also ensured that human experts remain in the loop. This uses AI-generating scenarios as decision support tools rather than as alternatives to local clinical decisions.
According to a Datarobot document, CVS Health reports that this AI-enabled approach will help run COVID-19 testing for over 29 million people and 59 million vaccine doses. The company continues to invest in the same predictive features to support a seasonal surge in demand, such as future public health campaigns and flu vaccinations.
The scale of this deployment spans more than 10,000 retail locations and highlights the complexity of logistics that AI managed during a critical period of national response. Although CVS does not disclose accurate ROI metrics, the ability to dynamically match staff and supply levels to move local demand was able to ensure increased resource availability and operational efficiency.
The system's enhanced intelligence framework allowed human leaders to maintain control while benefiting from AI-generated planning scenarios. This has been expanded to support the CVS's surge in seasonal vaccinations.
