Clinical intelligence in healthcare has advanced rapidly, with the global market valued at around $36 billion in 2025. Today, 90% of hospital systems have deployed some form of AI for medical imaging. Physician adoption doubled since 2023, with over 80% using AI in their practices. By almost every metric, artificial intelligence has arrived in clinical settings, and yet, according to Freddy del Barrio, founder of Companion AI, healthcare is still solving the wrong problem.
"AI in healthcare has been optimized for moments: diagnosis, alerts, predictions. But care doesn't happen in months; it happens over time," Freddy says. In his view, diagnostic models and imaging systems all have achieved precision at specific points in time, but the broader arc of care continues to unfold without a consistent layer of understanding.
His argument cuts to a structural flaw at the center of the prominent AI enthusiasm. The tools that dominate clinical settings today, while powerful, may operate episodically. According to Freddy, they activate at a single decision point and then recede, leaving no retained context. He argues that this absence of memory limits their ability to support clinicians consistently.
"The real problem isn't intelligence," he explains. "Healthcare has an abundance of data and models. The problem is continuity. No system is responsible for holding the full story of a patient over time."
He points to instances wherein a patient moves from a hospital to a rehabilitation facility to an assisted living center, where each record carries a history of preferences, behavioral patterns, emotional signals, and clinical changes. Within that context, Freddy argues that there is no cohesive system built to hold those fragmented points of data. "This diffusion of information prevents healthcare from forming a longitudinal understanding of the individual," he says. "That's the problem Companion is built to solve."
Companion positions itself as a persistent layer across care environments, one that is designed to track daily patient interactions, build longitudinal memory of behavior and preferences, surface context for clinical staff before every encounter, and keep families visible within the care loop. "The architecture is designed like a continuous presence. It doesn't ask what's wrong right now; it's asking what has changed over time," Freddy explains.
The healthcare industry, Freddy acknowledges, has started reaching toward more anticipatory models. In his view, predictive analytics, early warning systems, and population health monitoring all represent a genuine step beyond retrospective diagnosis. Yet he believes the industry is mischaracterizing what it's actually doing. Freddy states, "Most AI in healthcare today is reactive. Even what we call proactive is still event-driven. It triggers based on a threshold, a score, or a moment. What's missing is persistence."
Keeping reactive and proactive AI in the loop, Freddy highlights a third category: continuous, protective AI, which is always present. He notes that it learns longitudinally, supporting care across daily interactions. "Protective only works if the system actually works over time. Otherwise, it's just earlier reactivity," he says. Continuity, from his perspective, is what unlocks protective care, and without it, even the most sophisticated predictive models can remain incomplete.
The pressures accelerating this shift, Freddy notes, are converging in ways that make the timing urgent. Healthcare delivery is steadily moving beyond hospital walls, with home-based care models expanding alongside an aging population. In the US, adults aged 65 or older are projected to outnumber children by 2034, while workforce shortages continue to strain clinical capacity. Burnout among healthcare professionals remains elevated, with recent studies indicating that nearly 60% of physicians report symptoms of burnout across all 50 states.
In addition, social determinants of health, including loneliness and isolation, are becoming quantifiable clinical factors. Freddy connects these pressures to a misalignment between care delivery and technological infrastructure. "The system is being pushed out of the hospital," he says. "But the technology hasn't caught up yet."
Companion is built on the premise that it should.
Freddy emphasizes that adoption in healthcare depends on practical integration. Leadership within the company, he highlights, includes professionals with deep medtech and clinical experience, so that systems align with actual care workflows. Collaboration with operators forms a central component of product development. "This isn't being built in isolation," he adds. "We're building it with operators, not for them."
Internally, the emphasis on continuity is shared across the team. The company's Chief Technology Officer, Samuel Hootini, who has served across healthcare, real estate, e-commerce, and digital services, has approached the platform as a system capable of holding context over time. "Samuel didn't approach it like building another feature model; he approached it like building a system that holds memory," Freddy says.
On the operational side, Chief Operating Officer John Klepper brings leadership spanning healthcare and medtech, which Freddy considers essential to ensuring the platform works within actual care workflows. Brian Gewirtz, Head of Sales, works directly with senior living operators, the people responsible for deploying and sustaining care infrastructure on the ground. "Healthcare doesn't adopt ideas, it adopts systems that fit into how care is delivered," Freddy states.
Woven throughout Freddy's framework is a consistent position on the relationship between AI and clinical staff. Companion is designed to support caregivers, augment their capacity, and arm them with context that they would otherwise have to construct manually.
At a moment when clinician burnout is shaping policy conversations and staffing shortages are pushing entire care models toward restructuring, Freddy sees this augmentation function as central to the platform's value. He remarks, "AI that remembers means clinicians who spend less time rebuilding context and more time delivering care."
ⓒ 2026 TECHTIMES.com All rights reserved. Do not reproduce without permission.
