Dr. Andrew Marshall, a Meharry-trained emergency physician turned Google clinical specialist, is using artificial intelligence and digital health innovation to improve health literacy, expand equitable access to care, and help doctors deliver better outcomes worldwide.
What happens when an emergency physician with a computer science background steps into one of the world’s most influential technology companies? For Meharrian Dr. Andrew Marshall, it means reimagining how healthcare information can be delivered, accessed, and understood.
A 2015 graduate of Meharry School of Medicine, Marshall has built a career at the intersection of emergency medicine, bioinformatics, and artificial intelligence — using data and digital innovation to close gaps in care and improve health literacy worldwide.
His journey from the ER to Google reflects the evolving role of physicians in shaping the future of equitable healthcare.
A Nashville Beginning and a Problem-Solving Mindset
Marshall’s story, like many other Meharrian alumni, begins close to Meharry’s campus.
“I grew up in Nashville,” he recalls. “I remember coming to the dentist at Meharry with my mom when I was younger.”
Initially drawn to computer science — he attributes this to not being able to play video games as a child — he ultimately pivoted toward medicine, encouraged by his mother and shaped by a deeply personal experience.
When his father became ill during Marshall’s junior year of college, he began to see medicine differently.
“I remember sitting with him in the hospital at Vanderbilt and thinking about how certain things had been missed and his symptoms had kind of been ignored,” he shares.
Marshall’s “engineer brain” began firing. He took a step back and began to analyze the various points that got him there.
“Doctors know all this stuff,” he remembers thinking, “but then realizing that there were some problems still to solve.” His new focus became identifying inefficiencies and solving systemic problems in medicine — a feat he began at Meharry and has defined his career ever since.
Upon graduation, Marshall chose emergency medicine for its pace, complexity, and problem-solving demands.
He describes the emergency department as an environment where data flows rapidly and clinicians must think “algorithmically” — gathering information quickly and making decisive calls.
“I think about even my first shift as an intern and I was like, there’s just so many things going on, so many different types of patients, that you just start to see a wide variety of pathophysiology,” he recalls. “And then you also start to see where the hospital system fails a little bit.”
“I feel like the ER is the gateway to the hospital,” he says. “So it’s almost like a safety net. And I think what appealed to me about that is being there for people when they really needed me.”
In the ER, he witnessed firsthand where hospital systems fall short — patients lacking primary care access, specialty referrals, or flexible schedules. Those observations fueled his interest in digital solutions.
Bioinformatics and the Challenge of Digital Health
After residency, Marshall pursued advanced training at Harvard, blending biology, statistics, and computer science.
“Bioinformatics,” he describes, is the convergence of biology, statistics and computer science. “In the middle of that Venn diagram is bioinformatics.”
He later completed a fellowship in clinical informatics — a field that applies data science directly to patient care. This training gave him the tools to analyze large datasets, design algorithms, and build digital health systems from concept to implementation.
“It gave me a framework to approach the use of large data sets. Gave me a bit of a statistical background to be able to analyze my own data,” he says. “To take ideas from the start all the way to the finish.”
The challenge, Marshall found, was despite its promise, digital health innovation faces systemic barriers.
“One of the biggest challenges to innovation and developing effective digital health solutions,” Marshall claims, “is that our healthcare system right now is very expensive. And adding these solutions, even though sometimes they improve outcomes, can add more expense without increases in reimbursement.”
Marshall emphasizes that innovation in healthcare often feels incremental — “innovating on the margins” — but small improvements can meaningfully enhance patient experiences, and for him, it’s enough to keep innovating.
Designing Digital Tools That Improve Access
When evaluating digital health tools aimed at improving equity, once more, Marshall turns to the algorithmic approach, consuming large amounts of data to reach a “decision point.”
“Automated screening algorithms like the ones that I’ve explored and helped develop, they do the same thing where it basically helps you make a decision.”
But Marshall is quick to point out that while these algorithms may help clinicians make structured decisions, poorly designed ones can unintentionally reinforce bias — especially if they fail to account for geography, language barriers, or demographic differences.
“If you’re not building stuff with language multilingual support built in, then you’re missing a large part of the vulnerable population,” he says.
Equitable design must account for technological access and social determinants alike.
“There’s no one size fits all solution,” he states. “If you’re building something for a group of people in rural settings, you have to go and sit with them and see what they’re dealing with.”
Artificial Intelligence and Health Literacy
Now at Google, Marshall focuses on health information quality. With more than 500 million health-related searches daily, accurate information delivery can directly impact public health.
“If you can point people to the right information at that point when they make a search, then you have improved their health,” Marshall offers. “Because the internet is full of misinformation. And so just giving somebody the right information at the point when they make a search can really, really help them out a lot.”
He believes artificial intelligence can significantly improve health literacy — empowering patients to better understand diagnoses, medications, lab results, and treatment options.
“I personally do not think that AI is going to replace doctors, but I think there are a lot of doctors that need more resources,” he says, highlighting rural areas in the U.S. and outside the country that may not have a specific expertise.
“They could use a helping hand in making good decisions for their patients.”
Marshall acknowledges that equitable AI requires addressing device and technology access, representation in datasets, and community trust.
“You’re not going to get those folks represented in your data unless you can convince them that it’s safe to accept healthcare and safe to share their data,” he stresses.
Transparency, privacy protections, and education about data rights are essential to closing representation gaps.
In fact, this episode’s myth addresses the sobering reality that only 1-2% of global genomic data comes from people of African descent. Increasing representation in research is essential to developing equitable treatments and preventive strategies.
Marshall’s Message to Meharrians
Marshall’s advice to Meharry students and young health professionals alike is clear.
“Get familiar with AI,” he says. “Get used to it because it’s here to stay.”
He also emphasizes that healthcare professionals must hone their clinical skills first — then leverage technology as a tool for empowerment.
“Focus on being the best health professional you can be during your journey,” he urges. “You’re going to see a problem and you’re going to be like, I want to solve this problem. And from that, you can start to think about innovation.”
“It’s not going to replace you if you know how to use it,” he says. “It’s going to empower you to be a superhero.”
Be sure to follow Public Service Announcement with Dr. James E.K. Hildreth on Apple, Spotify, YouTube, or wherever you get your favorite podcasts. For more content like this, join the conversation online at mmc.edu/podcast.