When Stefan Bogdanovic walked into Google X in November 2019, there was no lab waiting for him. Just a near empty room, an early-stage idea, and a deadline.
His task: build the company’s first magnetocardiography (MCG) system, a tool to detect the faint magnetic fields produced by the human heart using diamond-based quantum sensors. The assignment fit his background perfectly. Stefan had spent years studying the quantum properties of diamonds, earning a reputation for precision, persistence and building things from the ground up. What he didn’t know yet was just how much of that persistence he’d need.
Jack Hidary, SandboxAQ’s CEO, laid out the first challenge. “He was hoping to see the new MCG system in a hospital setting by the end of the year,” Stefan recalled. “It was exactly one month and 10 days away. There was no lab at the time, and we were essentially starting from scratch.” To Stefan, it felt impossible, but he committed to making as much progress as he could in that time.
Starting from scratch wasn’t new to him. At age 12, a summer school lecture first hooked him on quantum physics. “I heard some crazy stories about quantum physics that made no sense to me, but I was hooked immediately.” That spark carried him from Belgrade to a PhD in the Netherlands, where he built his first lab with nothing but “an empty room and a credit card.” Later, as a postdoc at Harvard, he said the biggest lesson wasn’t technical at all: “I learned the difference between when things will work if you keep pushing, and when they will not work and you should save time by stopping early. Failing early gives you some perspective.”
That perspective was soon tested, when COVID-19 shut down Google X’s labs just months into the project. Stefan realized he didn’t want to repeat the solitude of his PhD. “I didn’t want to be alone in the lab again,” he said. So he called an old colleague, Christian Nguyen, who packed up his life and drove across the country with his partner and their cat to join Stefan. “And just like that,” Stefan remembered, “Our lab doubled in size!”
The two pushed forward, and were able to measure the first-ever human MCG signal in the lab using the diamond quantum sensor, something that has never been done before. But their first real test came during the first clinical study of their system at UCSF. The results were disappointing.
It was a painful but clarifying moment. “The great (and the terrible) thing of building a sensor with high sensitivity is that they will pick up everything, including the artifacts you aren’t intending to measure,” he explained. “The magnetic interference at the hospital was thousands of times greater than the human signal we were trying to capture.” That early failure turned out to be a gift. “Doing the UCSF study early on was maybe the smartest thing we have done. This taught us that the diamond quantum sensors we were developing would not do the job we intended them to do. We were too attached to the technology and had a ‘sensor-first approach’ which was proven to be wrong.”
For Stefan, it was also personal. “I was heartbroken because I thought that I was hired to develop a diamond quantum sensor.” He had a bit of an existential crisis and asked himself, “Am I a diamond quantum physicist, or am I something more?”
That question opened the door to a pivot. The team began testing other sensing technologies and landed on atomic vapor cell sensors, which could be scaled into multi-channel arrays that are very effective in combating environmental noise. Their return to UCSF with the new system was a success. "We were able to measure the human heart in the hospital, and the results we got were pretty compelling," Stefan said.
That shift reframed the whole effort. “Before that switch, I thought we were the ‘diamond quantum sensor team.’ After that switch, we became a medical device team. If you focus on the outcome you want to measure, then you’re not emotionally attached to the methodology. That allows you to switch tools when they don’t fit your needs and always choose the best tool for the task at hand.”
The challenges now are less about physics and more about producing clinical impact, health-economic outcomes, and clearing regulatory hurdles. The current generation of quantum sensors is still expensive, and scaling them means competing with established, cheaper tools like ultrasound and ECGs. “You might build the best, most perfect sensor system, and still no one uses it because hospital systems can’t get reimbursed for it. That’s the most difficult part,” Stefan said.
Still, he’s motivated by the highly personal impact that MCG could make on patients. “Initially, I didn’t want to do applied physics or develop products. I just wanted to stay in academia,” he said. “But now that I’m here, I don’t think I’m ever going back. In the clinical studies we are currently running at Mt. Sinai and the Mayo Clinic, we saw many case studies of how MCG could potentially address the current gap in standard of care tools. We still need to prove it but once you taste that impact…it’s a pretty great feeling.”
At SandboxAQ, Stefan’s adaptability has become a core part of his work. It’s what kept the project moving forward when the original plan failed. And it’s what may one day take magnetocardiography from an experiment in a lab to a tool that changes how heart disease is diagnosed around the world.
