HealthTech Execs Are Using AI Safely, Breaking Through Internal Resistance, and Accelerating Adoption

The C-suite has always governed technology at arm's length. When .NET replaced Visual Basic, a CTO's job was to understand the value, allocate the budget, and direct the team. AI has closed that distance. It puts analytical and creative capability directly in executives' hands, making hands-on experimentation the new prerequisite for enterprise adoption and productivity gains.

Shammy Narayanan is the Vice President of Platform Engineering, Data, and AI at healthcare tech provider Welldoc. A TEDx speaker, published author, and 11x multi-cloud certified engineer, he has held senior leadership roles at global healthcare and financial services firms including HCL Technologies, with deep expertise in AI and data strategy. He believes the AI era is shifting executive leadership from delegation to demonstration.

"AI is the first technology wave where leaders can't stay one layer removed. If executives aren't using the tools themselves, the organization will struggle to adopt them," says Narayanan. For previous technology waves, that distance was manageable through subject matter experts. Today, leaders can retrieve data, generate drafts, and visualize information themselves. The expertise bar for doing so has never been lower.

Relying on a specialist to utilize the tools is becoming a thing of the past, and the implications run deeper than convenience. As AI tools handle tasks that once required a dedicated request, an email, or a ticket, the traditional division between executive decision-making and technical execution is being redrawn. For those who don't make the shift, the competitive cost is real and compounding.

• From ticket to tool: "In the past, if I wanted a few data points, I had to send an email to my data team or raise a Jira ticket. Today, a leader can get that data themselves without needing to understand even the syntax of SQL. Fundamentally, you are less dependent," says Narayanan. As a leader, he notes, the same applies to everyday tasks: dictating to a tool that drafts a communication, or dropping text into a platform that returns a visual. The dependency on experts has sharply diminished.
• The cost of hesitation: In a landscape where AI capabilities are compressing decision timelines, the once-cautious "wait-and-watch" approach is a direct path to obsolescence. "The decision-making for executives is tricky. If I invest in hardware and it's obsolete in six months, and a leader is unfamiliar with what an LLM is or what tokens are, the team won't be able to take that bold step," says Narayanan. "That hesitation opens the door for your competition. If leaders don't talk the language of technology and still speak only in high-level English, it is a clear sign that the company is not headed in the right direction."

The shift in leadership behavior matters most where resistance is strongest. Narayanan notes this is often most pronounced among experienced staff, who may not embrace AI tools as quickly as their junior colleagues. Organizations are responding with a mix of carrot-and-stick approaches: some linking AI adoption to promotions, others making it a condition of employment.

• Seniority slowdown: "The adoption rate is high at a junior level, but people who have been there for more than ten or fifteen years have a passive resistance to AI. They rely on their own instincts, code from scratch, and their productivity is very low, only using AI when it's extremely difficult to proceed," says Narayanan. The gap compounds over time, quietly limiting output as junior colleagues pull ahead.
• Show, don't tell: Executive experimentation, he says, is the key to breaking through. "When a CTO does some 'vibe coding' and says, 'You know what? This is an app that I deployed. I just sat over a weekend and did a DIY project,' it sends a strong message even to the seniors that they better start to fall in line. The message percolates very, very fast," notes Narayanan.

But achieving cultural adoption often reveals a second challenge: pilot purgatory. Many organizations find themselves in a frustrating cycle of running pilots and experiments that never translate into production value. Without a clear organizational road map that defines the problem, teams are left directionless, Narayanan says.

• Needle in a haystack: "If you don't have a clear direction and you just say there are a lot of shiny tools available, 'let's try it out,' then you are basically trying to search for a needle in a haystack. You wouldn't find it," says Narayanan. The answer isn't a better algorithm or smarter code. It starts with a clearer problem statement.
• Business before bots: For Narayanan, escaping pilot purgatory requires a specific discipline about who gets involved first. "Always the product management team. They know what this feature can do, cannot do, how we want to differentiate. It starts there. And subsequently, once it is getting converted into user stories, you get the AI team involved," he explains.

The leaders who succeed in the AI era will share two instincts: a genuine appetite for continuous learning, and the willingness to abandon tools, vendors, or strategies that no longer serve the mission, Narayanan says. Perhaps his most pointed advice, however, is a human one. Rather than rushing to hire an entirely new AI team, organizations should look inward first and empower internal teams. "You need to be respectful of the team in-house. Having them in the mix ensures a lot of success because they know the blind spots, the black holes, and the nuances of the organization. Much of that tribal knowledge is a great lever for success. Organizations that don't understand this make costlier experiments and, ultimately, they fail," he concludes. "Once we get that foundation right, the superstructure will build on itself."