As personal “elevator pitches” go, Noosheen Hashemi’s is pretty darn compelling: “My passion is to close the gap between what is and what can be,” she writes at the top of her professional profile on LinkedIn.
Hashemi has been doing just that for decades, including her 10 years at Oracle (she joined the tech giant the year before its 1986 IPO), where she earned the nickname “Noosheen the Machine” from colleagues and formed a close bond with hard-charging Oracle co-founder Larry Ellison. She helped grow the company from $25 million to $3 billion in revenues and is credited with helping scale the operational infrastructure of one of the most productive and fast-growing sales forces in corporate history.
Since then, she has raised two children, run two foundations, served on a long list of boards, and invested in more than 100 companies along with her husband, Zod Nazem, through their family investment enterprise, HAND Capital. “My experience is vast because I have done so many things in my life, and I have lived every phase on steroids,” she acknowledges of her exhaustive resume.
These many paths began to converge in 2013, when Hashemi attended a conference on machine learning (ML) and its impact on society held at Stanford University, where she received her master’s degree in management in 1993. One of the speakers, Stanford computer science professor and prolific AI researcher Fei-Fei Li, discussed the intersection of ML and health. When Hashemi expressed interest, Li told her about a conference the following week on that very topic in Los Angeles. She promptly bought a plane ticket. “It’s really beautiful how the universe works,” she says, describing the spark that led to her latest ambitious venture: an AI health tech startup based in Menlo Park called January.ai.
January.ai founder Noosheen Hashemi (center) in the lab with the January.ai team
“The $3 trillion healthcare business is actually sick care; it’s all about decline, disease, and death,” Hashemi laments. “In addition, healthcare is not a market, there is a lot of opacity, and people have few choices, so consumers lack the agency that they enjoy in other sectors like stock trading, music, travel, what have you.” Rather than tackle the problem through policy and government, though, Hashemi set out to use market mechanism to empower consumers.
“The generic advice to solve multivariable problems right now is: Go lose 15 pounds and all your markers will improve. But most people can’t just drop weight like that and they get frustrated. ML and other technologies will allow us to solve multivariable health problems, and that is what I wake up for every morning.”
Machine learning, Hashemi believes, has the opportunity to fill in crucial blanks in the science of keeping people healthy. “We can gain more insights into our previous research, for one thing, but also redesign our research trials for the future to answer questions we were never able to answer before,” she says. Continuous monitors to track heart rate and blood sugar are changing the game, as wearable devices move beyond counting steps to helping manage our health.
January.ai has a beta product in the Google Play Store, Sugar.ai, that provides glycemic load of foods and predicted glycemic curves. January.ai’s machine learning has expanded the existing list of thousands of foods into the millions, and its goal is to eventually cover all foods. In addition to providing the predicted glycemic response to various foods among the healthy population versus those with prediabetes, the app breaks new ground by offering a window into each user’s personalized glycemic response, as part of a study the company is conducting called The Sugar Challenge.
“As it turns out, our blood glucose responses vary by person, and understanding one’s curves is not something that anyone delivers right now,” Hashemi says. “While people can see their glucose levels in apps, they are not connected to other data sources and streams and sources.”
January.ai’s technology combines many data modalities to understand its users’ physiologies and quantify their lifestyles, the combination of which Hashemi and her colleagues call Integrative Physiotyping. They have started with glucose and heart rate data; in time, they plan to add microbiome and classic markers such as cholesterol and triglycerides. “The generic advice to solve multivariable problems right now is: Go lose 15 pounds and all your markers will improve. But most people can’t just drop weight like that and they get frustrated. ML and other technologies will allow us to solve multivariable health problems, and that is what I wake up for every morning,” Hashemi says excitedly.
To take on these challenges — in largely uncharted territory — Hashemi, a business scaler, has surrounded herself with a diverse, multidisciplinary team of scientists, doctors, and engineers, she notes, starting with her co-founders, Mike Snyder, MD, and Justin Sonnenburg, PhD. Snyder has led Stanford’s Department of Genetics and its Center for Genomics and Personalized Medicine since 2009, while Sonnenburg is an associate professor of microbiology and immunology at Stanford. January.ai has a 22-person staff at present, as well as a Science Advisory Board that includes experts in microbiome, genetics, prevention medicine, multiomic integration, early detection, and immunology. Seven of the employees focus specifically on machine learning and artificial intelligence.
“The most important building block of a brand new startup is the team,” Hashemi says of founding her first for-profit company (she previously co-founded the HAND Foundation, a private family foundation, and founded the PARSA Community Foundation to encourage U.S.-based philanthropy among the Iranian-American diaspora). “You wouldn’t believe the amount of evangelism I had to do to get some of the best and brightest. Hiring a machine-learning team in Silicon Valley is very hard because there’s no history of reinforcement learning being applied to health the way it has been for autonomous vehicles, robotics, finance, and gaming. Also, we’re next door to Apple, Microsoft, Netflix, and Google.”
Of course, Hashemi brings significant value to the January.ai table herself. “The biggest asset is knowing what building a thriving business looks like — in culture, in work ethic, in business strategy, in speed of execution,” she says of her expansive experience in both the corporate and nonprofit realms. “Coaching founders reignited my passion for being a direct operator,” she continues. “Once you have tasted the adrenaline of fast growth, you’re always looking for it … So I decided to roll up my sleeves in a much deeper way and start my own company!”
January.ai takes its name from the month of resolutions, Hashemi explains — the time “when everybody says, ‘I’m going to change, I’m going to start again.’” The company’s logo features a calendar sign “to depict the importance of spending one’s life well,” she says.
“The idea of picking something really challenging to do and then starting from a clean sheet of paper…I guess that’s what I’m most proud of so far.”
Not surprisingly, coming from someone as multilayered as Hashemi, the choice of name goes even deeper; the word “January,” in fact, stems from ancient Roman god Janus, symbol of beginnings, transitions, passages, and also endings. He is most commonly depicted with two faces, one looking forward and the other looking back. “We think of the face facing back as the health cards you were dealt, such as genetics or the foods your family fed you when you were young,” Hashemi relates. “The face forward is a symbol of agency and reason — knowing oneself and making decisions to live one’s best.”
Hashemi herself seems to look only in one direction. “I singularly focus on one thing at a time, dedicating 7-10 years or more to bringing a vision to fruition. I can see my way through the next decade right now and it’s super exciting.” January.ai will come out of stealth with its upcoming MVP product, an iOS app that provides new insights around food as well as a personalized health report. In the coming years, the company’s goal is to form deep phenotyping of its users; in other words, helping them better understand their bodies and lifestyles based on their sensor data and personal information logs, information pulled from electronic health records, genetics, and more. “We want to help people figure out how to finetune their bodies and meet their health goals,” Hashemi says.
Asked to reflect on these early days of running her own company, Hashemi pauses for the first time in our lengthy conversation. “Am I grateful? Absolutely,” she says of January.ai’s promising start. “Am I amazed by the quality of our scientists and engineers? Absolutely. But I’ve never been the kind of person who stops and celebrates too much.”
She takes another brief pause before concluding, “The idea of picking something really challenging to do and then starting from a clean sheet of paper…I guess that’s what I’m most proud of so far.”