Argomenti trattati
The pharmacy and health services company CVS Health is investing in digital health tools and platforms that let design teams iterate faster while staying grounded in the lived experience of patients. Through a collaboration with Simile, CVS Health has created a set of agentic twins—AI-driven stand-ins trained on human inputs—that act as testbeds for messaging, workflows and new service concepts. Those simulated personas are calibrated using a substantial dataset so teams can see how representative users might react before rolling changes into care delivery.
At the core of the effort are millions of survey and behavioral inputs: the twins reflect responses aggregated from more than 400,000 people and include roughly 2.9 million consented responses across over 200 distinct behavioral scenarios. That scale enables CVS to use the digital agents for rapid experimentation and to explore paths that would otherwise be lengthy or costly to study with live participants, especially when targeting underserved or difficult-to-reach groups.
Building accurate simulations and why that matters
CVS’s approach begins with a structured intake process. The team conducts a moderated interview with participants to gather demographic markers and lifestyle information—things like diet, routines and typical health behaviors—so simulated profiles closely map to real-world customers. By capturing those attributes up front, the company can generate profiles that behave consistently across scenarios and act as reliable proxies during product development.
What the simulations reveal about patient needs
Using the simulated respondents, CVS design teams can rapidly test hypotheses about user experience and service nudges. The models surface patterns that matter to clinicians and pharmacy teams, and they do so without waiting for long-term field trials. In practice this means the company can iterate on communications, refill flows and outreach programs while observing which changes most influence actual behavior.
Medication adherence and actionable barriers
One clear use case is exploring drivers of medication adherence. The digital twins helped the team identify which obstacles are most likely to derail a treatment plan. Across simulated scenarios, issues such as trust in the pharmacy, confidence in how medications are handled, and the convenience of refill pickup or delivery surfaced repeatedly. Those insights allowed CVS to prioritize fixes—like streamlining refill steps—that directly address friction points and could boost initiation and persistence on prescribed regimens.
Trust, convenience and the clinician connection
The simulations also showed that patients value a personal link to their care team, especially pharmacists, and that perceived convenience strongly influences whether someone starts or continues therapy. Where a process feels stressful or cumbersome, some users indicated they might avoid beginning a treatment altogether. By simplifying workflows and enhancing clinician touchpoints, CVS has a pathway to increase both uptake and retention of treatments.
Validation, governance and keeping the patient voice central
While artificial intelligence accelerates discovery, CVS makes clear that the twins are not a substitute for direct human engagement; rather, they are an amplifier of the customer perspective. When a simulated response seems anomalous or unexpected, the team follows up with targeted research, which can include additional interviews or pilot studies. This layered approach maintains rigor while benefiting from the speed of AI-driven experimentation.
Governance and ethical safeguards are embedded in the process. Development activities operate under controls that monitor tone, fairness and safety so that outputs remain respectful and equitable. According to Sri Narasimhan, vice president and head of enterprise customer experience and insights at CVS, the goal is to center the consumer in every decision: the twins deliver near real-time, actionable feedback that helps the company design services people will actually use, while still validating and governing results through human-led checks.
In short, the partnership with Simile gives CVS Health a practical, data-rich way to simulate patient behavior, shorten research cycles and surface specific, fixable barriers to care. By combining digital twins with ongoing validation and clear governance, the company aims to amplify the patient voice and translate those insights into more accessible, trustworthy and convenient care experiences.