During the COVID-19 pandemic, hospitals across the country were operating under extreme strain. Physician burnout was rising as patient volumes surged, and inefficiencies that had long been tolerated suddenly carried immediate consequences.
While tools and technologies already existed to address many of these operational challenges, particularly around clinical documentation, the nature of healthcare complicates when and how new tech is introduced. For Nick F. Hernandez, Chief Technical Officer at ZyDoc Medical Transcription, the crisis exposed a fundamental constraint on innovation in healthcare. Even when proven technology promises relief, the act of introducing it carries risk, one that is often borne first by clinicians and, ultimately, by patients.
“Timing is often treated as a technical or operational concern when it’s really an ethical one,” Hernandez says. “By waiting for the perfect moment, we can unintentionally preserve systems that are already failing the people who rely on them.”
To support healthcare professionals during the pandemic, ZyDoc, an early stage company focused on clinical documentation and workflow efficiency, offered its technology at no cost to help reduce the documentation burden consuming scarce clinical time. “This wasn’t about product adoption or revenue,” Hernandez says. “It was about whether proven technology could meaningfully help clinicians care for more patients during a crisis.”
Many hospitals chose to continue with existing systems rather than introduce new tools in the midst of the crisis. An understandable and operationally defensible decision, it proved instructive for Hernandez. It clarified his view that effective technical leadership in healthcare requires anticipating how new technology will land, and who will feel its impact first.
“I don’t just ask, is this technically sound or is this low risk?” he says. “I ask, what is the cost of doing nothing and who bears that cost?”
Drawing the Line Between Automation and Judgment
That question has become increasingly urgent as healthcare organizations consider the benefits and risks of automation. Under pressure to do more with less, artificial intelligence is often framed as a solution in itself. Hernandez sees this framing as overly simplistic, and potentially risky.
“The hardest decisions aren’t about what can be automated,” he says. “They’re about what should be automated.” AI systems are effective at reducing cognitive load and handling repetitive tasks, but they do not understand context, accountability, or ethical consequence in the way humans do. In software development and clinical support alike, impressive AI outputs can obscure small errors that compound into systemic risk if left unchecked.
“AI should reduce cognitive load and administrative burden,” Hernandez says, “but not replace human judgment where accountability, ethics, and trust are essential.”
Preserving the Talent Pipeline
The push toward automation has also created a less visible problem. As organizations rely more heavily on senior professionals augmented by AI, junior and entry-level roles are disappearing.
“We’re increasingly relying on senior talent augmented by AI, while eliminating the roles that create the future senior talent,” Hernandez says. “That’s not just a workforce problem. It’s a safety and resilience problem.” Responsible innovation means designing systems that preserve learning pathways.
Healthcare systems depend on judgment developed over time and exposure to real-world complexity. AI can accelerate experienced professionals, but it cannot replace the process of learning or understanding why systems behave the way they do. Without deliberate investment in human development, organizations risk becoming productive in the short term while hollowing out expertise that is critical to both the profession and society’s well-being.
Innovation that Survives Reality
Delivering innovation under regulatory and clinical pressures demands the very best from health tech providers. Hernandez relies on three practices to balance innovation with safety, rigor, and respect for the clinical environments where technology is deployed.
One is resisting over-engineering. In healthcare, designing for every edge case can stall progress before clinicians ever see value. Hernandez favors constrained early versions that clearly block unsafe behavior and move forward, with the understanding that systems can evolve as real-world needs emerge.
“Don’t let perfection be the enemy of the good,” he says. “Clinicians need value now, not theoretical completeness months later.”
Another best practice is to pay close attention to who is not speaking. Engineers and clinicians often see risks early, but do not always feel safe or invited to raise them. Creating space for those voices helps surface blind spots that are easily overlooked in day-to-day workflows.
Hernandez also builds teams with the expectation that complex problems don’t have a single right answer. By encouraging disagreement early and pressure-testing ideas from multiple angles, teams surface trade-offs quickly and avoid false certainty before committing to a path forward.
The CTO as System Steward
As technology cycles accelerate, Hernandez sees the CTO role shifting from chief architect to system steward. Tools change constantly, teams are compressed, and expectations expand. The risk is optimizing for immediate productivity while eroding the foundations that sustain long-term resilience.
“AI can accelerate experienced professionals, but it doesn’t replace learning, judgment building, or understanding why systems behave the way they do,” Hernandez says.
Effective technical leaders will recognize that preserving human capability is now a core responsibility. Organizations that invest in judgment, context, and ethical reasoning will remain resilient as tools evolve. Those that do not may find themselves with highly automated systems and no one left who truly understands them.
Follow Nick Hernandez on LinkedIn for more insights.
