Burnout is rising, but doctors aren’t just asking for automation — they want tools that actually understand their work.
getty
The hospital pager beeps. The patient backlog grows. And somewhere between charting, coding and clicking through drop-down menus, the doctor wonders — when did this job become so joyless?
The healthcare burnout crisis isn’t new, but it’s growing a lot worse. Research by the AMA and Stanford Medicine in 2024 found that more than 60% of U.S. doctors said they were burnt out. The main reason was because they had to do too much paperwork and handled too many administrative responsibilities. The World Health Organization called burnout an “occupational phenomenon” in 2019, but it’s becoming more and more common in hospitals.
In response, several startups have come up with solutions, many of which use AI. These solutions claim to make taking notes easier, improve timetables and free up doctors’ time. But even if that goal is good, AI tools in healthcare have often caused just as many problems as they say they will fix.
Can these technologies truly help ease the pressure, or are they just making things more complicated in a system that is already strained?
The AI-Powered Turning Point
One company now trying to rewrite that narrative is Pieces Technologies, a Dallas-based health tech firm that recently launched a mobile conversational AI assistant aimed at clinical documentation — a pain point widely seen as a key driver of physician burnout.
Dr. Ruben Amarasingham, CEO of Pieces and a former hospitalist himself, says the new tool is designed to “operate almost like a bedside scribe in your pocket.”
The assistant listens, transcribes and generates medical notes in real time. It can also integrate directly with electronic health records (EHR), queue orders and adapt its interface to the clinician’s style. But what makes the product stand out, according to Amarasingham, is the company’s focus on designing for transparency and trust — qualities he believes are essential to “truly support clinicians.”
AI That “Assists, Not Impress”
The company’s newest release, Amarasingham says, is not just a product update — it’s a signal of a broader shift in philosophy across the industry.
“There’s a fundamental difference between building AI to impress and building AI to assist,” he told me. “We wanted this tool to disappear into the background — to feel like it’s part of the workflow, not something bolted on top of it.”
That way of thinking is similar to what a lot of people in the field are starting to push for: tools that put context, clarity and clinical relevance ahead of general intelligence.
Instead of using traditional training data, Pieces’ assistant draws from a proprietary process that includes prompting tailored to clinical tasks, a knowledge graph, and a comprehension engine designed for high-trust environments. In other words, it’s built to think more like a doctor — not just a chatbot trained on random data. It also offers granular control to physicians, allowing them to review and revise before anything is entered into the record. And most critically, Amarasingham noted, it’s designed to restore time — not just optimize it.
“In our early pilots, doctors told us they were finishing their days earlier. But more importantly, they said they were thinking more clearly,” he noted. “That’s the kind of impact that builds trust.”
Beyond Hospitals
The challenge of burnout, of course, extends far beyond hospitals. It’s a structural issue that cuts across industries, affecting everyone from teachers to tech workers. Sammy Rubin, CEO of the UK-based wellbeing platform YuLife, believes the medical profession has simply become the most visible casualty.
“Emotional exhaustion, long hours and high responsibility, which lead to burnout in clinical settings, are no longer exclusive to hospitals,” Rubin noted. “Whether someone is in a hospital or behind a laptop, the symptoms and impact on mental health can be just as serious.”
YuLife, which has partnerships with Bupa and MetLife, has built its model around proactive wellbeing — encouraging small, achievable actions like walking or breathwork through gamification and nudges.
For Rubin, the lesson for healthcare is simple: if support feels unreachable, it doesn’t work.
“Solutions need to respect time and energy,” he said. “They should work in the background and build over time, rather than demanding a complete lifestyle shift in someone’s most stressful moments.”
It’s a philosophy that resonates with the latest crop of clinician-focused AI tools — systems that are moving away from flashy dashboards and toward frictionless, human-centered design.
Designing With Purpose
Still, for many doctors, burnout is not just about time. It’s about meaning — the feeling that their work has become transactional, disconnected from purpose. But Rubin believes technology can help restore that, but only if it’s intentional.
“If a tool helps someone feel more in control, more effective, or more recognised for their efforts, it creates space for meaning to return,” he said. “Purpose can’t be forced, but the right conditions can help it grow.”
That means moving beyond features. For Amarasingham, it means building AI that behaves more like a colleague than a machine.
“We’re not trying to replace clinical judgment — we’re trying to respect it,” he enthused. “Burnout isn’t solved by reducing clicks alone. It’s solved when people feel seen, supported, and in control of their time again.”
It’s a high bar. But perhaps it’s also a necessary one. As healthcare systems edge closer to collapse, and the promises of AI continue to outpace their proofs, the question is no longer whether to build. It’s how to build with care and who we choose to build for.
