Neha Patki
I distinctly remember the night when I realized that I had a lifelong friend in Neha. It was 2018, I was walking down Market Street in San Francisco’s Financial District, and I was anxiously texting her.
The two of us were already good friends back then, but I had not yet gone through the existential quarter-life crisis of questioning the validity of one’s friendships, and so there I was, walking alone on a Friday night shortly after I had turned twenty five, and thinking about Neha and about all the ways in which she made that crisis feel less scary.
That particular Friday night, I went to a party where I felt unquestionably out of place. Maybe it was because I was still trying to figure out which San Francisco’s social scene to fit in, but whatever the reason was, I definitely was the odd one out talking about Jessie Ware’s concert in the midst of a Flip Cup game with strangers and a frenzied debate over what the correct way was to wear a toga if you were a guy (the consensus was that it counts only if you see the nipple.)
I knew Neha was home doing a wine-and-paint night with our friend Stephanie, and so I texted her, saying that it was a code-red moment and that I needed to escape. Just like she always does in these situations, she texted “Hi Denny!” and told me to come over whenever. I left the party faster than the speed of light and momentarily became a slightly maturer adult.
It’s that same sense of safety that I feel on the sunny but windy July afternoon in San Francisco when I leave my apartment to go meet Neha, who is visiting California from Boston, where she now lives and co-leads DataCebo, a synthetic data tech company. She moved to Boston two years ago, and ever since she’s left the Bay Area, San Francisco hasn’t really been the same for me.
We are at Duboce Park Cafe, our usual meetup spot in the city, and are drinking kale smoothies. Neha is here because she is visiting her parents in San Jose before she goes to Japan with two friends. I can tell that she is visibly excited about the trip, and when Neha is excited about traveling, I know two things have happened. Details of the trip have been ironed out and she has had time to research the place she’s visiting. Diligent research that most people would never do.
“So, one thing I learned that I had no idea about,” she tells me, “is that Kyoto was apparently completely spared from the atomic bomb in World War II.”
As is usually the case with me and history, I am clueless, but I don’t say anything.
Instead, I diligently write down Neha’s findings in my notepad. She tells me that she has spent some time over the last few weeks studying the three Japanese writing systems: hiragana, katakana, and kanji. She also tells me that hiragana and katakana are both phonetic, and that she actually finds them very intuitive, especially because the word order is similar to that in Marathi, which she can read. And then I laugh, completely in awe, when she tells me that she also prepared an inspo deck for the trip.
“Oh, and Denny, I don’t know if I told you,” she adds, “but I also made a new friend in Boston.”
Making friends as an adult is a topic that Neha and I have overanalyzed many times, so I obviously want to know what her winning formula is.
“Through work?” I ask.
“Actually, no. Bumble BFF.”
The dating-app-turned-friend-finding-app worked out for Neha, and I am not surprised by this because Neha is the person who will give any method a chance and then evaluate in a structured, analytical way if it’s the right method for her.
A few years ago, she wanted to become better at enjoying solo time in nature, so she took a book to the park, sat under a tree, and tried to do the whole solo-in-nature thing. The weekend after, she gave me a detailed cost-benefit analysis, very rightfully arguing that the cost of sitting on firm ground and waging a war against bees, while being too hot and then too cold and then too hot again, is too high and outweighs any benefit of being in the nature.
This time though, the cost-benefit analysis showed Bumble BFF was the right method for her.
“I think it’s easier than dating,” Neha says. “One, I think dating comes with a lot of expectations. Two, dating has a bunch of rules, and that’s what I felt especially here in SF, where it was all about the checklists and the metrics when you’re going on dates, just very mechanical. BFF is more organic, and meeting people as friends actually helped me understand how to do dating better.”
This is exactly why I feel that sense of safety when I am around Neha. She is a deep-thinker, someone who approaches everything in life with a very mathy mindset, whether she’s solving convoluted machine learning problems or figuring out the right watercolor palette for her art. She is nerdy, and charmingly nerdy, to be exact, because her curiosity is unconfined. She wants to understand why things are the way they are. And that means she is also very interested in people and why they are the way they are, which really means she always sees you for who you are, not for who you think you should be.
Another reason I feel safe with her is that Neha is unknowingly non-stereotypical. She is a woman in tech, she is nerdy and bookish, and yet she gets equally excited about watching Gilmore Girls together, helping me pick the right shade of foundation, or figuring out our outfits for a night out. She will get lost in the world of equations and numbers, and will later be creating a beautiful, captivating painting of the sunset. She is sharp and witty and intimidatingly smart, and yet she is also warm and kind to others. She is successful and ambitious, but she is not afraid to share her feelings when she feels lonely or anxious.
In other words, Neha is authentic. She means what she says and she speaks up when she sees something that doesn’t feel right. As we’re walking from Duboce Park Cafe to my apartment in Lower Haight, we get into a debate over Susan Sontag’s Against Interpretation essay and the validity of meaning in art, and she tells me about the time when high-school Neha “lost it” and openly questioned her English teacher’s excessive scrutiny of F. Scott Fitzgerald’s The Great Gatsby.
“Why did F. Scott Fitzgerald pick this color? What is the meaning of this color?”
I laugh as Neha waves her hand sassily, reenacting her high-school self.
“I was so sick of it, I just went ‘Well, what if F. Scott Fitzgerald just had a color wheel, spun the color wheel, and that’s how he landed on that color?’”
This eternal search for truth is at the core of Neha’s authenticity. Which explains probably why she was always a math person, and why she always lights up when she’s faced with a problem that requires numerical gymnastics.
Though she doesn’t remember this, my first deeper interaction with Neha was on a Wednesday night, when we were college freshmen in 2011, in our dorm kitchen, and when I asked Neha to help me out with a physics problem. I would fail the intro classical mechanics exam tomorrow, and I knew I was going to fail it, because that Wednesday night I kept staring at the free-body diagrams and the numerous equation variables with sheer horror, all while watching Neha patiently and gleefully explain her way of deriving those equations for me.
“I mean, I know it’s a quote from Mean Girls,” she says as we enter my building, “but what Cady said is exactly why I like math. It’s the same in every country. It just makes sense.”
After we finished undergrad, both of us stayed at MIT to complete our Master’s degrees. We became neighbors, and while I do remember Neha mentioning her research in machine learning to me, I didn’t have the knowledge—or maturity even—to understand what she was doing. We both later moved to San Francisco, and because Neha went on to build a career in product management, I think I forgot that she was still a programmer, and a math nerd, at heart.
But I definitely do remember the night in San Francisco when she reminded me what a math wiz she still was, the night she said she would leave San Francisco and move to Boston to join her grad school supervisor and build out a synthetic data company. A stealth company initially that would later be named DataCebo.
“How is DataCebo these days?” I ask her.
“DataCebo has been great,” she says. “The work tickles my mind in the right way. And I think I found the sweet spot. I love being in this space because my degree was in machine learning, but I also like the business and operational angle, which I learned through working at Google and YouTube. Plus, I really like that I get to build a company.”
The idea for the company came out of Neha’s research during her Master’s program. It all started when her advisor realized that he had to jump through so many hoops to enable his students to work with real-world data, which is something that’s undoubtedly essential for machine learning. The hoops were justified though; access to real and potentially sensitive data is something that should be restricted, made available only to those who truly need it. Otherwise all sorts of troubles can happen, like data leaks or misuse of personal information.
The question therefore was: could there be a way to mimic the datasets of interest and make them available for general use without actually exposing the data? That’s what Neha wanted to answer with her research. She built a system that automatically created synthetic data—called the Synthetic Data Vault—by building generative models of relational databases. Synthetic in this context meant the data points in these “fake” datasets were not the actual data points, but they still had the same mathematical relationships of the data points in the “real” datasets. These synthetic datasets also retained the format of the original datasets, from column types to missing column values.
Neha had essentially developed a tool for synthetic data generation and showed, by working with data scientists, that they could do the same data modeling and produce the same results when using the synthetic datasets instead of the real ones. And that tool became the foundation of what is now DataCebo. The concept itself is already impressive, but I have been even more amazed by Neha’s ability to go beyond her product leadership role, when I heard about the many late nights that she had spent setting up the company’s content and PR strategy. It was Neha who inspired me to start digitally drawing out ideas for my own analytics work, when I saw how she had designed eye-catching statistics graphs for DataCebo’s blog.
Again, this doesn’t come as a surprise. Neha is an artist after all. While we’re chatting, I show her the painting of a gorgeous burning sunset, hung on my kitchen wall, that she made for me a while ago. I mention that her use of colors, especially of orange and blue, is something that I have always loved about her art.
“The choice of those colors is actually very mathy,” she laughs. “When you look at nature, nature is far more desatured in terms of color. It doesn’t really have these super duper vibrant colors. So, when I paint, what I am trying to do is consciously desaturate. Blue and orange mixed together produce gray, so that’s the desaturation you get when you look at a painting like this.”
For Neha, painting is a way of seeing things. When she observes objects in space, she will often see them not just as objects, but as relationships between form and color. She calls this “getting stuck in art mode,” and it’s funny that she calls it art mode, because what it really sounds like is getting stuck in math mode.
“I do think painting is far more analytical than people think,” she says. “You have to see what you’re painting, and you have to see it for what it is. Not what you think it should be.”
As is the case with everything in her life, if Neha says something, she means it. And she means it when she says that she is conscious about her art. She even applies color theory to her sartorial choices.
“Color theory is very important to me when it comes to fashion. Bright colors just wash me out. Any bright clothing combined with the greenness and oliveness of my skin hue makes me very gray and desaturated.”
Another good anecdote of Neha’s committed personality is probably the night she tried to organize a wine-and-paint night with a few other girls in San Francisco. She prepared the wine and the colors and the art supplies, hoping for a cozy night of talking about everyone’s techniques as beautiful paintings got created. But that didn’t quite work out.
“The next thing I know,” she says, “everyone is drunk out of their mind and laughing and chatting, except for Neha, who is sitting there in the corner by herself, focused on the painting and trying to get the colors right.”
Neha has to take the train back to San Jose soon, so we go out for a walk and head over to Smitten on Valencia Street for some ice cream. We also want to stop by Dog Eared Books to check out if there is anything worth reading, primarily because Neha is a big reader. Funnily, she actually struggled with spelling for a really long time and even thought there was an “r” in the word salad. But her love of reading developed nonetheless, and by the time I met her in college, she had already read an impressive, Goodreads-worthy portfolio of books. I ask her if she thinks there is a formula for which books she ends up liking.
“No, not really. I think the right book has to find you at the right time in your life. A book that you read now might only click a few years down the road, when it’s the right time for that book. The right time for you.”
But she does have a few favorites. Neha’s top five books are:
Anxious People by Fredrik Backman
Pachinko by Min Jin Lee
Trick Mirror by Jia Tolentino
The Power of Myth by Joseph Campbell
Gödel, Escher, Bach: an Eternal Golden Braid by Douglas Hofstadter
The last two books are particularly interesting choices. The Power of Myth analyzes the role of myths in society, and more broadly the role of storytelling, which is something that Neha has thought about often ever since she took an AI seminar called The Human Intelligence Enterprise that was taught by Patrick Winston at MIT. Winston came up with the idea of the strong story hypothesis, which posits that human intelligence is different from that of other primates because humans can create, tell, understand, and repurpose stories.
Gödel, Escher, Bach: an Eternal Golden Braid is a non-fiction book, at an impressive length of almost 800 pages, that Neha read in her senior year of college. This book deals with the topics of consciousness, recursion, and intelligence in a format that is perfectly suited for Neha—puzzles. From what I can gather through Neha’s description of the book, it’s really about how we find meaning in the things around us.
At first I thought they were logical choices for Neha’s top five contenders because they deal with very mathy concepts, but after I thought about it, they are actually logical choices for Neha because, in a way, they seek to understand how we construct and present the truth of the world around us. Which is without a doubt the most Neha thing ever.
As we’re perusing the bookstore, I suggest that we buy each other a book that neither of us had read before. Neha would get me a nonfiction one, because I don’t like fiction that much, and I would get her a fiction one, because she doesn’t like nonfiction that much. She gets me a copy of Consider the Lobster and Other Essays by David Foster Wallace and I get her a copy of Girlfriend on Mars: A Novel by Deborah Willis.
“What should we do after we read them?” she asks me.
“How about each of us annotates the book they are reading, and when we are done, we switch and we mail the annotated books to each other? I read the book I bought you, with your annotations, and you read the book you bought me, with my annotations. Say by September 31st?”
Neha loves the idea. She says we have a deal.
We go grab our Smitten ice creams, and I realize that I can’t just propose this idea and forget about it. I have to read David Foster Wallace’s essays and annotate them, because when Neha says it’s a deal, then it’s a deal. And that means I will definitely be getting a copy of Deborah Willis’s book with meticulous annotations in the mail.
