Curing broken healthcare systems with AI: Zayed Yasin, WRITER industry lead

As medical costs continue to spiral out of control and patient satisfaction plummets, it’s clear that the status quo is no longer sustainable. That’s why this week on Humans of AI, we’re sitting down with Zayed Yasin, the healthcare industry leader at WRITER and emergency room physician on a mission to revolutionize how we approach healthcare.

In this episode, Zayed shares his vision for a future where AI enhances, rather than replaces, the human touch in medicine. From improving clinical workflows to closing the gap between patients and healthcare providers, Zayed shows us that with the right approach, AI can be a powerful tool for driving real change in the healthcare system.

From the emergency room to the AI revolution

When Zayed looks at today’s healthcare system, he sees a system that is fundamentally broken. Despite being one of the most technologically advanced industries in the world, healthcare still struggles with paperwork, bureaucracy, and inefficiency.

“Doctors aren’t happy, patients aren’t happy, nobody’s happy with what’s going on today,” Zayed says. “There’s some part of it that has to do with what we fund and how do we fund it? And those are huge political questions that technology can’t help.”

But, Zayed believes that technology — particularly generative AI — has the potential to revolutionize healthcare by automating the scut work that takes up so much of healthcare professionals’ time.

And he should know. Before joining WRITER, Zayed spent years working in academic emergency departments, where he saw firsthand the toll that administrative tasks took on doctors and patients.

“It just felt like a hamster wheel,” he recalls. “I wanted to do something that was going to be able to have an impact on a bigger picture level.” That’s when Zayed discovered generative AI while working for a value-based care organization. After building AI solutions for the care teams, the technology’s potential captivated him. He realized he wanted to focus on it more directly.

So Zayed reached out to an old college friend — who happened to be May Habib, CEO of WRITER. By late 2023, WRITER had dipped its toes into healthcare and realized two key things — the industry was a great fit for the company and its technology, but it required a specialized team and approach.

“I was a healthcare person looking for a platform,” Zayed says. “They were a platform looking for a healthcare person, and we took it from there.”

Where AI can truly add value in healthcare

AI integration promises significant improvements, but the stakes couldn’t be higher. Critics rightfully worry about putting life-and-death decisions in the hands of AI systems that have generated convincing but false information. The last thing anyone wants is a misdiagnosis or denied coverage stemming from an AI hallucination.

But Zayed argues that this fear-first approach is preventing the industry from addressing genuinely broken systems. “People have a tendency to put everything in the high-risk bucket and say, ‘This is too scary. Or let’s just wait,'” he says.

High-value, low-risk use cases

Zayed argues that healthcare organizations should take a more nuanced approach, identifying areas where AI can add value without putting patients at risk. For example, WRITER is working with healthcare company CirrusMD to develop a decision support navigator that helps doctors connect patients with services they may be eligible for.

Integrating a decision support navigator directly into the point of care allows them to analyze real-time doctor-patient interactions to instantly surface relevant benefits without interrupting the workflow. The results create a win-win-win scenario — doctors appreciate the streamlined experience, patients receive better care, and employers and payors see increased utilization of the benefits they’re already funding.

The push-button myth

On the opposite end of the spectrum lies another dangerous misconception that AI represents push-button magic capable of replacing entire workforces.

“For some people, that’s really exciting because it helps their budgets,” Zayed says. “For some people, that’s really scary because it’s their livelihood.” Both reactions miss the mark.

In reality, most healthcare AI applications require humans in the loop — and not just as passive observers. Technology must make human jobs easier, not eliminate them entirely.

Keeping humans in the loop

Zayed’s work with pharmaceutical companies illustrates this principle perfectly. His team develops AI systems for writing reimbursement dossiers — complex medical documents spanning 10 to hundreds of pages.

“It’s not that hard to build an agent that writes a paper or a dossier that’s, you know, 80 to 97% accurate off the cuff,” Zayed explains. “However, that’s not remotely good enough. But you’re building two applications for everything that you do.”

One AI system generates content, while another serves as a checker. Meanwhile, human experts use specialized tools to verify references and data — transforming what would normally be a tedious quality assurance process into something smooth and efficient.

Building AI-ready organizations from within

While identifying AI’s technical capabilities is straightforward — anyone can read the McKinsey and BCG reports — the harder question is whether organizations can ‌ handle the change required. Simply layering AI onto existing jobs provides marginal benefits at best. Zayed says that transformation starts when individuals fundamentally shift their roles — from writers to editors, for example.

“Following that is changing the relationships within an organization,” he explains. “Not just the rules of the individuals, but the way like medical affairs interacts with the commercial division of an organization or the way that legal and regulatory interface, those inter-organization changes are where I think the biggest opportunities are.”

Few enterprises have reached this level of organizational maturity. Most struggle just to implement change within individual departments, let alone tackle the cross-functional collaboration that requires strong C-level sponsorship. But according to Zayed, the transformation will be both bigger and slower than most expect. The defining factor will be mindset.

“This transformation is coming, and people have to ask themselves, either as individuals and as organizations, do they want to take an abundance mindset and say, ‘This is gonna make everything better, and I had to figure out how,’” Zayed explains. “Or a scarcity mindset, like ‘I need to avoid change for as long as possible.’”

Approaching AI strategy with a vision, not with fear

Organizations need to understand that the technology is unlike anything they’ve seen before. Zayed says AI has different performance characteristics that render old risk management tools and heuristics ineffective.

“You do need to be cautious when you’re getting started with something that just doesn’t work the same way other things work,” he advises. But it’s still important to start with a big vision and ask yourself, “How do I take the first step towards that big vision?”

Vision-driven organizations build momentum through continuous improvement, while fear-driven organizations spend months processing even successful results because they lack strategic direction.

“There’s no ‘AI strategy’ and ‘strategy,’” Zayed clarifies. “There’s just what’s your strategy and how can AI help you with that?”

Solving the personalized medicine problem

Modern medicine has achieved what Zayed describes as genuine magic — gene therapies that can cure lifelong chronic diseases like cancer, spinal muscular atrophy, and sickle cell disease. But this scientific breakthrough has created an unexpected business challenge reshaping how the healthcare industry operates.

“What they can do is miraculous, but you’ve gotta find those patients and find the specific doctors who are prescribing, treating those diseases, because not everybody can manage it,” Zayed explains. “These are very sub-specialized, complicated treatments. [The problem is] being able to find both of those people and then educate them and walk them through a very complex administrative and medical decision-making process to figure out ‘How do we use these magic drugs?’”

This is precisely where AI can create value. The technology excels at the deep customization and thorough analysis that these breakthrough treatments demand, an idea former HoAI guest Joanna Taylor spoke about previously. For the first time, healthcare organizations can be as comprehensive as they’ve always wanted —  systematically connecting patients with life-saving treatments instead of leaving these miraculous cures underutilized due to logistical barriers.

Closing the patient empowerment gap

So far, the conversation has centered on how healthcare and life science companies can use AI to improve internal processes. But what about the patient experience from a practicing MD’s perspective?

Let’s be honest, Google is a terrible doctor. But AI systems represent a significant improvement.

“They’re not good enough to replace your doctor,” Zayed cautions. “But they’re enough for people to expect a much higher level of bespoke explanation about what’s going on.”

Patients now expect much higher levels of personalized explanations about their health conditions because they can see what AI tools accomplish on their own. Meanwhile, healthcare systems continue operating with decade-old communication approaches, creating a stark gap between technological capability and the actual patient experience.

“What we are looking for at WRITER is the organizations that have the courage to — not find ways of saying no — but find ways of saying yes and saying like, you actually can make things a lot better for everyone by leveraging the technology,” he says.

AI can eliminate ‌barriers currently separating doctors from patients and the treatments people need. In Zayed’s ideal future, doctors effortlessly access everything they need to know about medications and patient needs, while patients easily find the right specialists without navigating information overload or bureaucratic maze.

The administrative work doesn’t disappear. It just gets handled by machines instead of people. This shift returns medicine to its fundamental purpose — creating great treatments, prescribing them effectively, and ensuring patients can use them with minimal red tape.

The organizations that will thrive in this new era are those brave enough to challenge their own sacred cows, asking not “how have we always done this?” but “how should we do this now?” In healthcare, where the stakes involve human lives and well-being, that courage to reimagine established practices isn’t just strategic — it’s essential.

Want to hear more stories from the humans working at the crossroads of business and generative AI? Subscribe to Humans of AI wherever you listen to podcasts.