MOUNTAIN VIEW, Calif. — Google’s brash attempt to revolutionize medicine as it did the Internet is facing turbulence, and many leaders who launched its life sciences startup have quit, STAT has found.

Former employees pointed to one overriding reason for the exodus from Verily Life Sciences: the challenge of working with CEO Andrew Conrad.

Verily, one of Google’s “moonshots,” pursues ambitious, even radical, ideas that could take years to pay off. The emerging Silicon Valley juggernaut has attracted elite scientists, engineers, and data crunchers, and inspired buzz about its futuristic projects — as well as envy among competitors nervously eyeing this upstart with a seemingly unlimited bankroll.

The three-year-old venture has operated largely out of public view and carefully manages its image; employees said talking to a reporter without permission is a firing offense.

But people who know Conrad or have worked with him said in interviews that Google has entrusted its life sciences initiative to a divisive and impulsive leader whose practices are driving off top talent and leaving openings for competitors. They said many employees in key jobs were dispirited, and described a lack of focus and clear priorities that is unusual even in the chaotic culture of startups.

STAT identified a dozen top Verily managers, scientists, and engineers who have departed in the last year. Some gave up coveted Verily spots to return to the Google mother ship, including Diane Tang — a revered “Google fellow,” the company’s highest technical rank, achieved by only about a dozen employees in the company’s history. Others joined competitors.

Verily continues to grow and attract new talent, and in the hypercompetitive environment of Silicon Valley, executives often jump ship in search of richer compensation, fresh challenges, and more authority.

But “if they are getting off the roller coaster before it gets to the first dip,” something looks seriously wrong, said Rob Enderle, a technology analyst who has tracked Google since its inception. Those who depart well-financed startups usually forsake potential financial windfalls down the line, which further suggests that the people leaving Verily “are losing confidence in the leadership,” he said.

No similar brain drain has occurred at Calico, another ambitious Google spinoff, which is focused on increasing the human lifespan.

Conrad and others at Verily declined repeated interview requests over the past 10 weeks, and the company subsequently declined to answer written questions that described STAT’s findings.

It is easy to see why Conrad, a scientist and entrepreneur, caught the attention of Google co-founder Sergey Brin. His off-the-charts ambition and elite connections in Washington and scientific circles match Google’s grand plans for Verily, which Brin last year called “a huge undertaking” that he hoped would “transform the way we detect, prevent, and manage disease.”

Brin and his cofounder, Larry Page, gave Conrad a commanding role at Verily, which has no outside members on its board and no scientific advisory group — unlike most biotechs.

One former employee joked that Verily should be called “AndyCo.”

People who know Conrad described his eloquence about science and ties to the rich and powerful. A 52-year-old, publicity-shy biologist, he sports floppy gray-blond hair, a bushy goatee, and informal attire that echo the surfer’s boyhood in Malibu.

Friends call him charming and farsighted — a “disruptor,” in the words of former Paramount Pictures head Sherry Lansing, who will “always challenge the way things are done.”

Dr. Elias Zerhouni, a former director of the National Institutes of Health, praised Conrad’s “bubbling imagination,” “enormous IQ,” and “childlike excitement,” which let him “explain very complex concepts with extraordinary simplicity and passion.” Zerhouni is now the research chief of Sanofi, a drug company collaborating with Verily.

Former employees, however, characterized Conrad in less complimentary tones. They said he exaggerates what Verily can deliver, launches big projects on a whim, and rashly diverts resources from prior commitments to the next hot idea that might bring in revenue. This has led to what they describe as difficult meetings with business partners, and resignations by demoralized engineers and scientists in the face of seemingly impossible demands.

Beyond headquarters, people who have worked with Conrad said he has strained Verily’s relationships with important regulators in Washington, including officials at the Food and Drug Administration and the Department of Health and Human Services.

STAT interviewed dozens of former colleagues and others who know Conrad or his work. They included six former Verily or Google employees and contractors, and six current or former collaborators. Many asked not to be named, saying they wanted to protect ongoing relationships with the companies. STAT also reviewed hundreds of pages of public records, and scientific, legal, and patent documents authored by or involving Conrad.

Two former employees said Conrad often leaped at business partners’ ideas in meetings, saying that the company was already working on the very thing — even though it was news to senior staff members. He would assign engineers and scientists to work up the ideas, which often proved infeasible.

Former employees, including a one-time manager, said Conrad kicked employees out of meetings if he deemed them too junior, even when they were leading the project under discussion.

Many Silicon Valley techies are work-obsessed and used to long hours, but Verily has become so demanding, erratic, and unforgiving an environment, a former employee said, that he regularly found himself at his desk at 11 p.m., in tears. Crying, he added, was not uncommon inside Verily offices and restrooms.

Several key Verily leaders in addition to Tang have returned to Google, including Kobus Jooste, one of Verily’s first employees and a leader of major projects; Michael Pearson, a top business manager; and highly-regarded engineer Karl Townsend.

Townsend decided to leave along with nearly every other member of the sizeable team that built Verily’s “connectivity bridge,” a product that connects medical devices to the cloud, another former employee said. Conrad suddenly lost interest in selling or further developing the technology, this person said, and while team members could have stayed, they worried about Conrad’s fickle leadership.

Dr. Mark Lee, a top Verily scientist, joined another biotech startup. Jean Wang, already one of Google’s most accomplished engineers before joining Verily, worked there for 14 months before she decamped for Amazon.com in 2014. Babak Parviz, an inventor of the Google Glass wearable computer, cofounded Verily’s project to create a contact lens that continuously measures glucose in tears before he also left for Amazon around the same time.

“Leaving a hot Google property for Amazon — you don’t see that happening a lot at the top levels,” Enderle said, because Google enterprises are known for paying well and offering superb work environments.

While current or former Verily employees wouldn’t speak openly, Michael Luther, former president of a North Carolina research institute Conrad helped found, was blunt. “We used to joke and call (Conrad) the ‘seagull of science.’ He used to fly in, squawk, crap over everything, and fly away,” said Luther, now a biotech consultant.

“You couldn’t engage him for more than 10 minutes,” Luther added. “It was sort of the overpromise, under-deliver.”

DEFEATING MOTHER NATURE

Conrad talks loftily about what Verily can accomplish by merging technology with medicine, delivering a sales pitch in his occasional public remarks that meshes with the expansive vision of Google.

In short, he wants “to defeat Mother Nature,” Conrad told STAT in a brief interview last fall when the Verily name was unveiled. He said the name reflects a quest to unearth deep truths about health and medicine by cross-fertilizing specialists from diverse fields — big data, engineering, computer science, and even philosophy.

“We have to understand the ‘why’ of what people do,” Conrad said. “A philosopher might be as important as a chemist.”

Verily — spun off last year as a separate subsidiary under Google’s new corporate parent, Alphabet — provides few specifics about what it’s up to. Set up as a limited liability company, it does not have to disclose the names or duties of board members, and has not. The spawn of a Web titan didn’t have a website until December.

Potential hires were given only vague ideas about what they would work on until they were on board, a former manager said. And its researchers have published just one scholarly article — about a spoon for people with hand tremors.

But Verily, which employs about 400 people, has dribbled out some selected details. Its initial projects include “Baseline,” a study to track clinical, molecular, imaging, genetic, and microbiome data from 10,000 people over five years to learn what it means to be healthy — as the starting point for identifying the earliest signs of cancer and heart disease.

It is developing, with partner Novartis, a smart contact lens to continuously measure glucose levels in diabetics, and robotic surgical tools with Johnson & Johnson. Verily says it has two wristbands in the works: one to read diagnostic nanoparticles that would detect cancer or other ailments at very early stages, another to continuously monitor skin temperature, pulse, and other heart activity.

“A new car has up to 400 different sensors. You know the oil pressure. You know how much air’s in your tires. But we don’t do that for people,” Conrad says on a promotional video posted on Verily’s website. “Instead of episodic, reactive health care, we should provide preventative and proactive health care,” personalized to the individual.

Such ideas continue to inspire and attract high-profile talent, including Dr. Thomas R. Insel, a former director of the National Institute of Mental Health who joined Verily in November.

Massachusetts Institute of Technology professor Robert Langer, who serves on the board of Boston-based Entrega Bio, a Verily collaborator on the nanoparticle effort, called Verily “a fairly visionary group.” The work with Entrega has, so far, hit its milestones, Langer said, praising Verily’s fresh perspectives and Google’s big financial commitment to life sciences research.

He knew about some of the departing employees, and said Verily is “certainly moving very fast, so there may be growing pains.”

CAUTIONARY TALES

Conrad’s salesmanship was a big reason his first foray into the biotech business, helping to run a genetic-testing startup, was wildly successful. But some of the issues that are dogging him now first cropped up then — especially a tendency to alienate close associates.

Conrad was a moped-riding, torn-jeans-wearing graduate student at the University of California, Los Angeles in the late 1980s when he met Dr. Peter Schmid, a research professor. Schmid taught Conrad about molecular cloning and other laboratory techniques while he earned his PhD in anatomy and cell biology. They became friends, and together helped build Los Angeles-based National Genetics Institute, where Conrad served as chief scientific officer. The firm perfected efficient genetic tests for HIV and other ailments.

He had “a particular talent of talking nice and smooth,” Schmid said in a recent interview. “We were a good team. I invented this stuff and he sold this stuff.”

Richard Smith, a scientist who, as vice president of National Genetics, worked for Conrad for about a decade, called Schmid “a genius when it came to equipment,” and Conrad “the charismatic one-man marketing department.”

But conflicts emerged over apportioning credit for the company’s inventions, when Conrad and two others filed for three patents for improved testing technology without including Schmid as a co-inventor.

Schmid’s name was left off the patent applications “despite the fact that I had the idea, discussed it, and sent a fax with details to Andy,” he said in an email. He said the US Patent and Trademark Office added his name “after I was able to prove that the concepts originated from me.”

Patent office documents show that in each case, Schmid was added as a co-inventor after the original filings.

Conrad “had a tendency to try to push others out of the game to get ahead,” Schmid said. “He was afraid he would disappear in my shade.”

But Mike Aicher, CEO of National Genetics at the time and a friend of Conrad, credited Conrad and a collaborator from another company as the primary inventors. “To think that it was Peter’s idea — I would argue, absolutely not,” Aicher said.

Laboratory Corporation of America, or LabCorp, bought National Genetics in 2000 for $56 million, plus up to $16 million in incentive payments, according to Securities and Exchange Commission records, enriching Conrad and Schmid.

Both men stayed on with LabCorp, and after Conrad left the company in 2013, the two remained collaborators in a separate business venture involving human longevity.

Still in his 30s, Conrad was a multimillionaire — rich enough to catch the eye of David H. Murdock, the billionaire pineapple and real estate tycoon.

THE MURDOCK FACTOR

Conrad’s rise to future-of-medicine evangelist began in earnest at an art auction on the Hawaiian island of Lanai. Murdock, the Dole Food CEO who owned the entire island, was there, too, according to Luther, who knows both men. Conrad outbid Murdock.

Rather than be disappointed, Murdock saw in Conrad a kindred spirit who had enough money to not need anyone else’s.

“When you are a billionaire, every person who goes to see him, at the end of the day, wants something from him. He hated that,” said Luther. “David Murdock trusted very few people, but Andy was one of those.”

They were well-matched. Murdock, then a robust, late-70s mogul preoccupied with healthy living and longevity, needed a life sciences guru. Conrad needed a mentor.

Conrad had arrived in Lanai in 2000, when he built a lavish vacation home on a rise above Murdock’s Four Seasons resort golf course on the island’s breathtaking south coast. He was focused on his investments, leisure, and his wedding late that spring to Courtney Thorne-Smith, a star of the prime-time TV soap opera Melrose Place and the drama Ally McBeal.

The couple split early in 2001, but Conrad’s partnership with Murdock was blossoming.

Lynn Safrit, the retired president of Castle & Cooke, a real estate company owned by Murdock, said the Dole executive grew unnerved watching his bright but apparently unmotivated young neighbor “laying out in the sun, walking around with his swim trunks on.”

Murdock, a driven, self-made industrialist who left school in the ninth grade, advised Conrad to become more ambitious.

“I think he saw in Andrew … someone who could do great things,” said Safrit, who also knows Conrad well.

Conrad took Murdock’s advice to heart. Despite working for LabCorp, he made time for side projects with Murdock, becoming the billionaire’s jet-setting travel companion, scientific advisor, and confidant. In return, Murdock emerged as Conrad’s business patron, putting him on the boards of Castle & Cooke, Dole, and another firm he controlled, NovaRx. Conrad invested with Murdock and his son Justin.

His loyalty to Murdock had its limits, however. In 2013, Murdock moved to buy out other shareholders of the publicly-traded Dole. Conrad was the board member selected to manage the privatization. He and other directors concluded that Murdock’s price was too low, and “Murdock became furious,” wrote a judge in a ruling for a subsequent legal dispute.

In a voicemail quoted in court documents, Murdock told Conrad, “it sure as hell pisses me off to think that you didn’t call me and tell me what it is going on with you. I’m not accustomed to having a friend double-cross me, but if that has happened …” The recording ended there.

Murdock declined an interview request through a spokeswoman, saying that Conrad had asked him not to speak publicly about their relationship. (Conrad zealously guards his privacy: He has no Twitter account or Facebook page, and a Google search yields nothing linking him to his celebrity second wife, fashion model Haylynn Cohen.)

Ultimately, Murdock raised his price and Conrad’s committee signed off. Shareholders still sued, and last year were awarded $148 million in damages. Conrad was found to have acted properly, and was not held liable. When contacted by STAT, fellow directors said he acted with consummate integrity.

Before the run-in with Murdock, Conrad had gained valuable experience building and running the billionaire’s ambitious life sciences research enterprise — in a sort of practice run for Verily.

Murdock made Conrad the unpaid chief scientific advisor of the North Carolina Research Campus — a public-private venture on 350 acres near Charlotte that Murdock hoped would attract academic and industrial partners to conduct research and develop products focused on agriculture, food, nutrition, and health. The centerpiece was the David H. Murdock Research Institute, which they equipped with sophisticated scientific devices.

Conrad steered the project’s development and brought in big-name scientists, among them Dr. Robert Califf, now commissioner of the FDA, to join the institute’s board.

Conrad also encouraged Murdock to underwrite a precursor to Verily’s Baseline study, named after the billionaire and initially run by Califf at Duke University. In all, Murdock put about $1 billion into the enterprise, some of which he hoped to recoup through lease payments.

The campus opened amid the recession in 2008. As Conrad struggled to recruit businesses, he sometimes overpromised, said Luther, the Murdock Institute’s former president.

For example, he said Conrad told Murdock that he would get LabCorp to invest millions of dollars. The company set up a nearby repository to store and commercialize biological samples, but the big money never arrived.

Drawing a comparison to “the emperor has no clothes” fable, Luther said, “Andy kept telling Murdock … (what) he wanted to hear.”

As a managerial tutorial for Conrad, however, the venture was a big success, according to Luther. He said Conrad absorbed Murdock’s impatient approach to building a business.

Murdock would say, “Get your CEO on the phone and get him on the phone now!” Luther said. “Andy learned first-hand at the foot of the master.”

A HEAVYWEIGHT DEPARTS

Verily recently announced that it will soon trade the tranquil water features and shade trees of a satellite Google campus in Mountain View for a fortress-like cluster of buildings near the San Francisco airport that can accommodate a workforce of 1,000 — more than double the company’s current size.

But one person who won’t be joining Conrad and his expanding enterprise is Jeff Huber.

A heavyweight at Google X, the in-house incubator where Verily began as a project in 2013, Huber was an admired advisor to Conrad and other Verily leaders, regularly attending senior staff meetings. Highly regarded in technology circles, he lured top talent to the startup and helped land key corporate partners. Huber stayed at Google X after Verily was spun off as a separate Alphabet subsidiary last August, but he remained an integral, day-to-day force at the new firm.

Huber left in February to lead Grail, a startup that will try to find and cure cancer before symptoms show up.

Among Grail’s chief competitors is Verily.