Dr. Mylene Yao, Univfy

Anyone who has attempted to navigate the IVF process knows just how harrowing it is — physically, financially, and emotionally. Among the myriad frustrations is that doctors might estimate that a given cycle has, say, a 30% chance of succeeding, but he’s basing that largely on a finite set of experiences. As such, in the great majority of instances, it’s a guess.

Dr. Yao, a board-certified OB-GYN, was curious why some IVF patients had more success than others. While at Stanford conducting early embryo genetic research, Dr. Yao built a prediction model for IVF outcomes.

“Was it just luck, or did people go into it with different chances of success?” she asks. “It felt as if there was more to it than just age or general health. What we learned was that IVF could be more individualized to a woman’s healthcare profile.”

In 2010, Dr. Yao left Stanford to run Univfy, the company that sprang out of her research, full time as CEO. The question immediately shifted: Was the prediction model she created replicable elsewhere?

“A lot of doctors asked me, ‘Does this really have to be customized to my clinic? Couldn’t you just do a more generic prediction model, like an online calculator?’” Dr. Yao recalls. “We didn’t think that was good enough.”

Univfy partnered with myriad institutions to create doctor-­specific prediction models over the past few years, yet one unexpected challenge still remains: communication, especially between doctors and patients.

“We deliver information in a personalized patient report that doctors can use while talking to patients, but the probabilities associated with IVF sometimes aren’t heard the way they’re intended,” Dr. Yao notes. “A doctor might say there’s a 50% chance of success, but the patient hears that as there’s a 50–50 chance it won’t work.”