Developer Tea

Data Science w/ Elena Grewal (Part 2)

Episode Summary

In today's episode, I talk with Elena Grewal, head of data science at Airbnb. We cover a wide variety of topics, so make sure you catch the first part of this interview as well! Today's episode is sponsored by Fuse! Build native iOS and Android apps with less code and better collaboration. Head over to to learn more today!

Episode Notes

In today's episode, I talk with Elena Grewal, head of data science at Airbnb. We cover a wide variety of topics, so make sure you catch the first part of this interview as well!

Today's episode is sponsored by Fuse! Build native iOS and Android apps with less code and better collaboration. Head over to to learn more today!

New Promo Code: “dt” will give you listeners 70% off for 12 months. 70%!!! The code must be redeemed by December 31st 2017.

Episode Transcription

Because the reality is that being on your own and being self-directed is good, but only to a certain extent, right? Like you really want to be able to tie what you're doing back to a business problem at some point, or have a vision for how it will be useful and not stay purely in the theoretical forever. If you have a personality type like mine, then you're probably intrigued by academia. I really appreciate the academic environment, the ability to learn in a way that allows you to fail a low stress environment. But if you're also like me, then you would like to take that knowledge and use it. And Elena Graywall, that's exactly what she did in the most extreme way. She was deeply into the academic sphere. And then eventually she decided to go out to Airbnb. And if you heard the first episode, then you already know that the first part of this interview. And we're going to continue this discussion today. Elena is going to tell her story a little bit more in detail. Thank you so much for listening to today's episode of Developer Tea. My name is Jonathan Cutrell. And this show exists to help developers like you, whoever you are, wherever you're listening to help you uncover your career purpose. Hopefully something that Elena or I say on today's episode gives you some kind of spark, some way of seeing your career a little bit differently. Maybe growing the way that you perceive yourself or growing the way that you perceive your coworkers or your title, what you're doing in your job. Or maybe if you aren't a developer, perhaps it's going to spark an interest in development that you didn't have before. So that's the goal of this show. Thank you so much for listening. I'm going to get out of the way. We're going to get straight into this interview with Elena Graywall. Elena, thank you so much again for joining me for a second time for listeners who heard the first part of this. This is definitely the second time that actually it's technically the third time that we've tried to do this. We've had a few technical difficulties and you know, long story short. It's been a few weeks since I last talked to Elena, but thank you so much for taking time out of your day once again to meet with me and discuss a few things here on Developer Teana. Thank you, Jonathan. Well, it's great to be here and to talking with you and with your audience. So I'm excited to kind of pick up where we left off previously, but we are going to change gears a little bit. And we were talking about, you know, mentorship and having somebody who can kind of guide you through machine learning. We discussed your background and how you kind of accidentally ended up in this position because you were in a academic setting, but then that you wanted to apply that information outside of the academic setting. So you could actually see the fruits of your labor. Is that a relatively comprehensive explanation of your experience? That's perfect. Great. So I'd love to kind of rewind back. There's another experience that I want you to kind of talk about if you're willing to. You went and did an extended trip in India. And this is something that I personally have been really interested in. My wife and I are fascinated with Indian culture. In fact, we quite literally ordered Indian food tonight to eat. So I want to hear, you know, first of all about that cultural experience, but also a little bit more about why you decided to go to India. Yeah, yeah. Well, you're inspiring me. Now I kind of want to order some Indian food tonight. Great idea. Well, so, you know, my father is Indian actually. And grew up in India, grew up in Calcutta, and came to the US when he came for grad school. My mom is American. And it was interesting because growing up I hadn't really traveled to India. All of my family would come to stay with us in New Haven, Connecticut. And so I had experienced India through people coming from my family to visit and stay with my family. But I hadn't really spent much time there. And so, you know, my initial motivation was really to just go and to have that experience at staying with my family. I was in college and, you know, this summer after your first year in college is a great time to spend an extended period of time traveling. And I got a fellowship to go. And it was sort of related to, at that time, I was like, oh, I'm going to be a doctor. And so it's like, oh, I'll do something that's related to medicine. And I was going to work at a public health group at a medical college in Louisiana, which is where my family was living at the time, my uncle and aunt. And it was a research project. So it was actually related to data collection. And it enabled me to essentially accompany local medical practitioners to different parts of the city, even some of the rural areas. And it was just like amazing opportunity to understand what it was like to live in that place and to work there and to meet lots of different people. And, you know, one thing that I love when you land in India is that you truly feel like you're in a different world when you're there. The culture of the surroundings, it definitely, you're like, okay, this is this is a big world that we live in. And there are a lot of different places and, you know, different cultures. And that that was really fun and stimulating. And it gave me an appreciation of different cultures, which really has helped me to think about Airbnb in a way that, you know, I find really useful that like people might interact with our site differently in different places. And what would it mean to stay at an Airbnb in India, right? Like that's a totally different experience than staying in another culture. And it seems obvious, but I think like when you go and you spend an extended period of time, like that's something you really understand differently. And that's something that's so great about travel. And, you know, for me, it also was very transformative because I went to India thinking, oh, I'm going to be a doctor. And, you know, was doing this research project. And it was a very sexy topic. It was a cute, direal disease in children under five, which is one of the leading causes of death in children under five, which is terrible. Because, you know, the water isn't clean. There's unsanitary conditions. And, and children will get, you know, this illness and then they die from the dehydration. And so, you know, it's like this big problem. And I kind of realized that like, you know, I didn't want to be a doctor anymore. Because there's like, you know, I'm going to be on the receiving end of all of these problems with a government delivering services, you know, with these kind of systemic issues. And I don't want to be kind of on that end. I want to be on the end of like, you know, how do we fix that problem, you know, what's going on with public policy? Why is that the way it is? And, you know, that was one of the motivations for me to take a shift in my path away from medicine to kind of understanding the root causes and, you know, what might be changed. And that's where that kind of focus on impact comes from to. It's a really fascinating story. And it's interesting that you mentioned how I was actually the next question was going to be, you know, how does this impact the way you see you know, the world today and more importantly, well, maybe not more importantly, but similarly, how does it impact your work? And you mentioned something that actually reminded me of the time that actually came and visited Airbnb's office. I came out to San Francisco to hang out with the folks at spec and go to a conference out there and drop by Airbnb. I have a friend that works at Airbnb and walk through the building. And I saw some really cool stuff. Probably my favorite aspects of the office is, I guess it's not really an office. It's a pretty expansive building. Aside from the fact that the kitchen is really awesome, there's also all of these rooms that have been and maybe it's changed since I came, but I assume that they're still there. They're decorated like real Airbnb rooms that you that you can actually go and stay in. I am currently sitting in New South Wales which is a room. And so the idea is that we basically have listings and our homes that we pick and then one of the rooms is replicated for the conference room. It's a much better way of saying what I was trying to say. Listings is the word that I was looking for because that all encompassing term for, you know, a venue, I guess is. All the different types of places that you could stay. We like to use the word homes as well. Yeah. And that's such a cool thing because I bring that up because it reminds not only the people that are coming through the office, but it reminds that people who are working at Airbnb. This is really kind of the product. This is what your end point here is this experience. That's totally right. And that's what makes it really fun is to be close to that experience that you're creating and enabling people to go to these different places and have a new perspective that they might not have if they were staying in kind of a more cookie cutter living situation without that connection to the local place. So that's been really fun. And you know, I think the other thing that always gets me for Airbnb is just looking at where we have homes on Airbnb. That was something that continues to amaze me where, you know, I remember a friend of mine at Airbnb the first year I joined was like, well, I'm going to Mongolia. And I'm staying at Airbnb the whole time. And I was like, wait a minute. We have Airbnb's in Mongolia. And it's like, no, we don't just have one. We have a lot of Airbnb in Mongolia. How is this possible? Like, this is the coolest thing ever. Yeah. It is an amazing, amazing reality now. And really excited about, you know, this kind of direction for lots of services that have kind of taken that model, the concept of sharing and applied it across. I'd love to ask you, you know, working at Airbnb, especially in data science, you know, one of the things that I noticed, I actually came in and went to like an evening talk that was put on near the kitchen. That's why I knew about the kitchen. I actually had dinner there. And it was very cool. I was talking about different types of search and how we could optimize search and how Airbnb is optimizing search. So can you kind of detail maybe one or two or three, if you have them on hand, experiences that you've had recently or things that you've done recently that you think are, you know, really interesting, really exciting, maybe even kind of innovative stuff that you are involved in, that you think is just kind of energizing. Yeah. That's a great question. Oh my gosh, there are so many projects right now that are happening across the company. I mean, I think I mentioned, you know, we have 120 people on the team now and embedded in every single part of the company. So, you know, some of the cool things that we're working on really are so different and span so many different areas that it keeps it really fun. One area that I've been working on that has been really interesting to me is actually in our customer support field. So, you know, customer support is so important on Airbnb. You know, if you have a problem, oftentimes you're across the world, maybe you don't speak the language, there are lots of different things that can be going wrong. And so we have to be very fast and good in our response to when people are having a problem either using the site or when they're traveling. And the range of challenges that people can face is pretty large. Like, it's not a short list. Airbnb is not a simple product to use. And you know, there's a huge variety there in terms of like what can go wrong. And one of the challenges that we have is that, you know, when something goes wrong, it's not something that like any one person can address. Like, usually there's like a specialized person who can address that. And so, you know, we want to make sure that we're routing the issue to the right person. And so that's a really interesting case of how do we use the data that we're getting in about what the issue is, whether it's text data, whether it's voice data, and essentially understand that using machines to some degree and help to find who's the right person to answer this so that we can get that fast turn around time. And, you know, the idea is not to, you know, remove the human, but it's to assist in the process to help the customer support team to effectively respond and, you know, get the right kind of questions that they're prepared to answer quickly. And so that's a really interesting space. I mean, I think that's one of the cool things about machine learning and deep learning is thinking about how you can extract information from text or images and, you know, information that you really didn't have that same understanding of before those techniques were more readily available and widely applicable. So that's been really interesting. A lot of cool applications there. And really across all of our text data and image data, you know, previously you just have things like, well, how many photos are there? And that was a variable that you could use in your analysis. And now there's like more that you can use, which is so cool. And so that's been really interesting to see. And a part of that is also the automation of machine learning. So we've had a significant effort underway for probably the last year and a half around machine learning infrastructure and building out tooling so that, you know, any data scientist or engineer or product manager can say, hey, I think this this aspect of our product should have a model behind it. And it's really easy to get the data that you need to put it the model into production to test the model and to diagnose how it's working and how you can improve it after that. And so that's that's efforts that have been ongoing in the infrastructure world. And that's been a really cool project to see. And, you know, a piece of it is also the education project part of it. So, you know, how do people understand how to use machine learning and also data in general. So we have this awesome program called Data University that we launched this year probably at the beginning of year. We had a pilot before that. And essentially it was data education for anyone at the company. So we had data of 100 series, data 200 series, data 300 series, starting with, you know, how do you ask a good question? That's often the hardest thing for people to do is, you know, people will say, oh, I need this data. But then you're like, well, why do you need the data? What are you trying to answer? And so the first course is really about what is the heart of the question that you're trying to answer? And why does it matter? And that's what we start with. And then we go to helping people to be able to self serve and get the data they need to be able to visualize it in an effective way. And then, you know, ultimately to be even more advanced in order to say, hey, how do I use machine learning in my product? And so that's been a really cool thing that we've launched that has transformed the company in terms of how we're using data. Over a thousand employees have taken data university now. We have promotional video, which is hilarious. I never thought we would have a data-U promotional video. We have stickers. T-shirts, people wear the t-shirts all the time. And, you know, they seem like they're little things. But when you see that all over, it starts to create this just massive cultural shift in terms of how people are using data at the company. Yeah. So I'm looking at this this medium post about data university. And it's really it's really interesting. First of all, it reminds me quite a bit just to make it kind of a correlation in the design world. It reminds me a little bit of IDOs human-centered design. It's very much so around the human data. You know, the very top kind of assertion is that data is the voice of our customer, which we mentioned previously in the first part of our discussion. And it's really attractive design and everything. But the other part of this is that, you know, it's kind of like, I don't know, if you, neither of us were really in business probably in the 90s, I certainly wasn't. And so back then, you know, people were adopting computers for the first time in their businesses, especially small businesses. For the first time, they were using, you know, Microsoft Word. And, you know, it kind of felt like that was a big jump then. And that only those kind of geeky people were actually setting up computers in their offices. And then it became very much commonplace, right? It was, you know, by the end of the 90s, very much so commonplace. And it feels like this is kind of a future literacy that will be similar. This idea that, hey, you know what? It doesn't really matter if you're customer service or an engineer or a designer or, you know, an intern for that matter. Data is going to be important to you in these kinds of business initiatives. Yeah, no, I 100% agree. I mean, at the heart of it, it's really about honing your critical thinking skills and understanding how numbers and data can change what you're doing and inform what you're doing. And that's something that shapes everyone. I mean, you know, even going beyond that, I get really excited about it because, you know, something that I've seen is oftentimes we hear data and numbers in the news from our friends. And a lot of people don't know how to think about that, right? Like they don't know how to evaluate it to say, you know, does that make sense? Does it not make sense? How do I make my own conclusion? And so, one thing that I'm really excited about is promoting data literacy at Airbnb. But also giving people skills that, like you said, will last them in their life beyond Airbnb and serve them well in those other areas too. Yeah, yeah. And, you know, again, this isn't just Airbnb scale that that needs this information. There's so many other ways that even small business can benefit from this kind of thinking, you know, one of the things that I've started doing for myself, even in the small business world that we operate in at Whiteboard and in agency world, is I'm collecting quantitative data about what other people at the company think about me. And it seems so simple. But the idea is, you know, how, you know, which of these things would you categorize me as? And then I list a couple of categories of, you know, kind of roles that I want people to see me as or I want people to perceive me. And so the idea is, you know, if you start learning, that's kind of step one, right? You don't have to go through, you don't have to go through intense class. You don't have to figure out all of the math to be able to understand how to learn from data. If you start from that perspective, and then eventually you say, hey, you know what? As it turns out, there's this whole extra set of tools. There's this whole, you know, this whole field of statistics and then there's this whole field of, et cetera, et cetera. I've eventually getting to things like deep neural networks or whatever you, wherever you head in that direction. If you start by understanding, hey, there's information here in this data. And there's things that you can learn from it. There's ways that you can change, even at your very personal level, that whole, the whole quantified self movement, this can apply at a very personal level to help you understand what's going on rather than just guessing about what's going on. Yeah, and I mean, you know, we all have biases that we can recognize. And so, you know, that's where data can help us to say, you know, is this really what's going on or is it not? And I just kind of want it to be true. And those, the data really is kind of measuring, right? It doesn't really matter how wide I think my desk is. Thinking how wide my desk is, it's not really an effective thing for my brain to do. And you mentioned something earlier that I think is really important to kind of hone in on for a second. And that is the idea that this machine learning effort is not necessarily intended to replace humans, but instead to magnify our efforts, right? So to multiply our efforts. So the idea of, you know, getting someone to the right person quickly, this is a hugely valuable thing, right? Those people are not removed from their jobs. There's not a person that is, you know, the router, the customer service router. It's the customer service people that somebody reached them incorrectly and they have to reroute them, right? So that's not really a good use of their time as a human. As a human, a good use of their time is to practice empathy and to understand a unique circumstance. And, you know, for lack of a better term, pull some strings for the customer, right? That's the whole, the hope that a human on the other end is going to care about what I'm going through. Yeah, I mean, that's exactly right, that, you know, it's really about supplementing and assisting and giving us new opportunities. I mean, you know, sometimes people ask like, it was funny. I was in a panel the other day and someone was like, well, you know, we won't need data science anymore. It'll all be automated. It'll all be distributed. And I'm like, I don't think so. Like, I think there'll be, there'll be things for us to do. And it'll just be new things. And it'll just be a different type of thing. And, you know, even in the short five and a half years, I've been at Airbnb, you know, what I've worked on and the types of work that team has done has changed in part because we explicitly automate what we're doing as much as we can because we know that that's the way to scale. And it has not resulted in a shortage of work for us. We have just done the new work. Yeah. Yeah. Well, it's funny because we tend to even engineers, even once we know how this stuff works, we tend to believe that somehow everyone will collectively become better at programming. And that like collectively, we will make the, you know, better decisions about how we create this stuff. And that's not necessarily true. And a lot of what we as humans have to figure out how to do is decide what to learn, right? Decide in what direction are we going to point this thing? How are we, you know, what questions are we trying to answer? Those are the really long processes that a computer is going to be, first of all, completely inefficient at doing, right? But also the computers don't really have the contextual mind that a human does. We draw on experiences like from, you know, 15 years ago when I was a teenager, I'm drawing on an experience from then consciously or unconsciously to decide what to say next in this conversation, you know? So when we're designing systems, when we're designing things that are for other humans, those contexts are incredibly important to those decision making processes. Yep. No, it's definitely true. So that is where we can continue to explore and ask new questions and have a lot of fun. Today's episode is sponsored by Fuse. If you are interested in getting into mobile application development, then maybe you've actually come across Fuse. I'm going to go ahead and tell you it's come a long way. The industry has come a long way. And Fuse is pushing it forward that the very edge of where mobile application development is going. In fact, Fuse just went into 1.0. They're no longer in beta. It's a free tool that runs on macOS and on Windows. Now, if you are a mobile application developer, then you know this is true. Those tools that you've used for so many years, those the basic tools that you've always been using to develop mobile applications, they can be really clunky. They can be hard to understand. They can require a lot of code to do something very simple. And Fuse is trying to change that. They want you to be able to build applications with less code and more collaboration to be able to see the things that you're building a little bit easier. If you've ever used Unity for game development, then Fuse is essentially Unity for app development. There are tons of great examples available on And you can see all the source code for those examples as well. And for the super users, the people who know that you want to work in applications, you are a professional developer already. Fuse has a paid Fuse professional plan. It includes a built-in UI kit, which they call it UX kit because it also includes animation and navigation, premium charting, camera components, and with Fuse professional, you can actually use your Fuse projects inside of other projects with Xcode and Android Studio integrations. Most people don't need the paid plan. Most people can get by without the paid plan. But if you do decide to use a paid plan, you get the Fuse Studio, which is the editor that Fuse provides to the professional users. So you can get started for free today by having heading over to slash plans. slash plans. Now here's the thing. This is maybe the deepest discount that I've had on the show at any point. Fuse has provided a promo code and it expires at the end of this month, but the promo code is DT and you can get 70% off for 12 months, 70% off. That's not 30% off, so you're paying 70% of normal price. It's 70% off, so you're paying only a third of what it would normally cost less than a third actually. Go and check it out slash plans and use the code DT at checkout. Thanks again to Fuse for sponsoring today's episode and for providing a free tool. It's so incredibly useful for mobile app development. So I do want to shift slightly again. We've talked about your experience of going from wanting to be a doctor to the academic world and then eventually into the job that you're at now doing machine learning with Airbnb. I'd love for you to share if you don't mind maybe a dark moment or a difficult moment in your kind of path. At any point along that path, a dark moment that you felt like you didn't really see a way forward. Definitely. There have been many moments that have been low. I know it's punctuated by those. The time that comes to mind was, I guess it wasn't I didn't see a path forward, but it was a very stressful time. I guess I didn't see a path forward without a major change. It was when I took the job at Airbnb and I actually hadn't finished my dissertation. So at the end of your PhD, you have to write this long paper called a dissertation, which is like the culmination of all your work and it's meant to be research that's new and innovative in the field and it will be evaluated by your advisor and other professors for you to be able to get your doctorate. I hadn't finished mine and I was like, oh, you know, this will be fine. I'm just going to work on weekends on it and I'm going to have this full-time job and it'll be no problem. I definitely did not anticipate how much work starting at Airbnb would be having a new career and one that I was so excited about and got so into that it really did consume so much of my mental space that I had very little mental space to finish my dissertation. It was getting time to defend the dissertation and submit it and I just had a panic attack and I was like, I can't do this anymore. I'm too far behind. I'm not going to pass. I won't get my PhD. This is terrible. I've spent like six years working on it. How could I do this? And I went to my manager at Airbnb and I said, I need to take a month off of work. I have a month before I have to submit this. I can't do it anymore. Like, I have to take a month off. Can I just take like an unpaid leave or something like that? And thankfully, my manager must have realized how adult I was at the time and that it was just like not even an option and valued me as an employee. So they were like, okay, yeah, go ahead and take a month off. And so I ended up taking that month of unpaid leave from Airbnb and finishing my dissertation. And I was really glad that I did that. You know, it enabled me to have the space to do it well in the way that I really wanted to. It had been this combination of, you know, six years of work. And I just felt like I couldn't, I couldn't not do it and not finish it. So that was, that was really, really good that I did that. And but it was, it was a dark time. I was like, really stressed out. I was like having health problems. That was when I realized like, oh, like our bodies and our minds are connected. I mean, I knew that before. But I was like, oh, like, if I'm really, really stressed out, like, it'll impact everything. And so that was, that was really tough. And I was like, I don't know if I'll be able to do it. But, but I did. And it went really well. And then I was able to start back at Airbnb again. And, you know, I think that the quotation that came to mind when I thought about that time was Ruth Bitter Ginsburg has this great one where she says, you can have it all, but not at the same time. And so understanding like, you know, sometimes we think that we can do all the things, but, you know, maybe that will be tough to do. And we'll need to prioritize and think about how much time they'll actually take. And maybe at one point, this one thing's critical and the other point at the other thing is. And, you know, if we, if we kind of space it out in the right way, then then probably we can do all the things that we need to get done. But, um, you know, I was definitely one of those moments where I was like, oh, I bid off way more than I could show. But, yeah, there's another similar quote that says, you can have anything, but not everything. That's right. I can't remember who says that. A very similar, very similar kind of concept and staying focused. You know, that's a really interesting thing. And I was going to ask you, you know, what, what you feel like you learned from that experience. Certainly, you just mentioned something important that you learned that the mind and the body are definitely connected. No matter how much we want to, maybe sometimes deny that and try to soldier on that they are connected. What else would you say that you learned during that period? Yeah, I think just being a lot more realistic about my timelines for things and, you know, kind of, and that's something that you learn when you work too is, you know, you have to think about, okay, like here's this task that I need to get done. Let me break it down. You know, what do I need to do to finish it? How much time is that going to take and be realistic and, you know, leave buffer times and don't be kind of flippant about it where you're like, oh, it'll, it'll just get done. You know, I'll just do it on the weekends. Like, well, how much time will you spend on the weekend? Like, what does that really mean? And, and then prioritizing, you know, being a lot more careful about the kind of rank order of what you're doing and making sure that you don't end up in a situation where you're so overwhelmed because you really have taken on more than you can do. Yeah, it's so important to note kind of the action step that you took that was so critical. And that is the elimination step. This, this idea of focus is very often kind of abused in common discussion that you can focus on anything you want to. The truth is, focus requires elimination, taking things away and simplifying. And this, this can be at a very physical level, even, you know, simplifying your desk. This is actually a really common thing. If you Google, you know, how to, how to gain focus, you're going to see a bunch of listical articles and almost all of them say, clean your desk off. That's like one of the top. And there's some, some good science behind that. If you have a dirty house, even if you go to work, you leave the house behind, you still feel that clutter for whatever reason our minds carry this clutter around. But this is true in the philosophical sense or maybe in the, you know, kind of more ethereal sense that if you have a bunch of things that you are putting your effort into, like a bunch of different jobs that you do. For example, for a little while, I was doing soft skills weekly. This was probably about 10 weeks that I did it. And it was a weekly newsletter that I was, you know, putting a couple of articles together, which doesn't sound too difficult, right? It seems like, okay, you should be able to send a newsletter out. That shouldn't take too much time. The reality was the space that that occupied in my mind was far more than the task, you know, really required that kind of slot for projects that I'm currently doing that has a minimum size that ultimately was too much. And I decided to eliminate that so that I can make this podcast better. So I can make another side project called Beyond Boot Camp a little bit better. And by eliminating that thing, I gave myself, you know, I granted myself more space to do other things better. Yeah, no, that makes sense. And I mean, in that process too, you probably realized more to in terms of what was really important to you and and gave you that perspective to help to guide your decision about how to invest your time. Absolutely. And nothing will do that more than having your first child. And, you know, this is certainly not something that everybody's going to do in their life or wants to do for that matter. But for my wife and I, we had our first child this year. And that has been very much so clarifying in terms of priorities. And we very quickly learned what we actually value versus what we thought we would value. That makes a lot of sense. That's great. So excellent answer to such a difficult question because, you know, bringing up those things is not easy. But I do want to kind of go to the opposite end of the spectrum as well and ask you, was there a moment where you feel like you had kind of a life changing positive epiphany, something that stands out where you can look back and say, yes, at that moment in time, my mind changed. And if so, what was it about? Well, I would definitely say my trip to India was one of those. I talked about that one already about, you know, I'm just going to not be a doctor. I mean, I would say taking the job at Airbnb was one of those moments. You know, I think I mentioned to you no one in my program had ever done such a thing. My advisor was like, what are you talking about? You're going to be a data scientist. Like, we don't know what that is. My parents were really confused. They were like, what is the need? They're like, we don't know even know about this. And you're at Stanford getting a PhD like, we had high hopes for you. But now we're like kind of concerned what's going on. It was really interesting. And, you know, I think that like the epiphany that I had was just that, you know, I'm in data science, I'm a pretty rational person. But I would say that like probably of all the data scientists, we did like a Myers-Briggs test at one point for all of our data scientists. And there's this one that's like thinking versus feeling. It's the T versus the F. And I was the only F on the entire team. And I think back of that moment of deciding to come to Airbnb. And it was like, you know, there was a rational component to it, for sure. And I was looking at the data and I was looking at, you know, looking at Google trends, how many people are searching for Airbnb? What's the outlook here? Like getting as much data as I could. But at the end of the day, I was like, you know, this just feels right. Like it just feels like these are good people. They're smart people. I can't put a finger on the energy that I'm getting here to say like, this is the reason. But like, I know this is the right thing. And so that was I think a great moment for me in terms of like trusting myself and just taking a chance. And then kind of getting the feedback later that like actually that was a really good decision. And you know, kind of keep doing that. Like listen to listen to the data. But also think about like what are the other signals that your gut might be getting? That data that you can't put into a number, but it could help you to make a good decision. Yeah, that's that's such a good kind of directive for I also am an F. I'm an INFJ. What was your what were the rest of yours if you don't mind sharing? Oh, we had well, we're very close. I was an E and FJ. Okay. Yeah. And so this is there's there's a lot that can come out of this. I'm actually reading a book right now by Ray Dolly. I'm not sure if you've heard of him. He's an investor and he runs a company called Bridgewater. I think it may be called Bridgewater Associates. In any case, Ray Dolly has a book called Principles. He's actually going to release two volumes of this. And in the first volume, he talks about life principles and work principles. The other one's going to be more about finance and that kind of thing. But this the book has some really interesting things to to kind of wrap your mind around. He was actually one of the first financial firms to grab a hold of the idea of encoding their decision making process into code. So they created some decision making algorithms and they could plug in the numbers and effectively come out of the other side with what is it that we want to do? And he discusses some of that in the book, but he also discusses the importance of psychometrics. And he has everybody on his team take like three or four different psychometric tests and then he creates these baseball cards for each person. And these baseball cards kind of share people's strengths and weaknesses. It's a very interesting concept because I don't know that I would want my weaknesses presented on a baseball card per say at work. But this has been really effective for them because what they've been able to do is become very honest with each other and kind of form teams within bridge water. Once they get past that painful period of saying, okay, yes, I am actually not good at everything. Once they get past that, there's kind of this light bulb moment for them where they say, okay, I can actually collaborate with people who are better at certain aspects than I am at certain things than I am. So I thought it was interesting to even mention my breaks. But I also like the idea that you kind of went off of an impression even though you were also doing this other study, right? That you mentioned something in our previous discussion that maybe there's some data in your gut, right? Like maybe there's something in there that makes sense, but you can't quite rationalize. I 100% believe that at all times. And that's where I thought about, like when I walked into Airbnb, there was an energy in the air. You know, you could see it in the people around how they were moving, what the space was set up. And that's not a data point, right? But like that goes into your gut and you're like, wait, like I feel really good about this. Right. Yeah. Yeah. And you could probably quantify it with other things. But you know, in that moment that you're making that decision, you might not have exactly the numbers to say why you're deciding something. Yeah. Yeah. And it's difficult to know. I mean, a lot of the time we're wrong. And we have to learn how to, you know, pull ourselves out of that. Some of these epiphany moments like what you've experienced, you know, I had epiphany moments even at the very kind of basic level with, you know, JavaScript learning about objects and learning that I could inspect objects in the inspector. When I was first learning how to code, that totally changed the way that I thought about, you know, actually developing which JavaScript, because I was doing a lot more visual testing. And now I could actually do some comparing in the console. And this is brand new to me. I was, you know, as very young in terms of coding. I didn't know anything about what I was doing. And so I've had those kinds of moments where, you know, I realized, for example, that I've had the realization that almost every language that we use has some core fundamental concepts that are the same. And, you know, naming ends up being a pretty vast majority of what you do as a developer, right? Which is, you know, and that may not be as true. That's certainly true for web developers who are doing presentational information. I kind of think probably a little bit less true for you, though. Now they think about it because you're doing more math intensive things, but certainly you're still naming variables and that kind of stuff, right? No, that's very important. And also just thinking about how you can standardize, I mean, having a standard database and schema that is understandable is really important. So yeah, absolutely. So this has been an excellent discussion. I'd love to know, you know, there's a few things that I think everyone who listens to episodes like this, they are connecting to your big story, but also, you know, you have an everyday kind of process of operating as a data scientist. I'd love to know, you know, when you set out to learn something for yourself, particularly if you're looking to learn a skill, maybe learn a new language or understand a new concept, you know, what do you do to learn? What is the process that you walk through? I'm a Googler. That's my meaning. I think anybody who does something technical is like Google, yeah, it's like the best thing ever. So I mean, usually I will just say, hey, how do I learn how to do this? Type it into Google, see what pops up. I have a couple people that I look to as thought leaders in terms of how to learn something. So especially when it comes to deep learning and AI,, my friend Rachel Thomas and Jeremy Howard have some great blog posts about, you know, that area. And so, you know, I would look to some of those blog posts to link out to other blog posts that I could read to learn about the topic. But, you know, honestly, it's really Google. I've done courses online, things like that. But the other thing that I would say is whenever I'm trying to learn something, I tend to be a learner where if I'm not doing something that has a purpose, I have a really hard time fully mastering it. So I'll usually try and learn a skill in the context of I want to do this thing. And in order to do the thing, I need to learn the skill. And so, you know, if I want to learn a new skill, I'll think about, well, where could I apply that? And that, for me, makes it so much more real and also helps me to truly learn it because, you know, it's one thing to watch a course online and do some problem sets. And it's another to be like, okay, I need to solve this. How do I code this to solve it? Yeah, yeah. That's a really good point. And I think most people are probably very similar that learning, you know, in a hands-on way, especially when it comes, you know, right now I'm actually working on a react native project. It's one of the first that I've ever done. And I'm learning about all these various concepts that I haven't really been using in other frameworks and that kind of thing. And there's a lot to, there's a lot to grasp there. And if you've never done it, you know, it's even harder to grasp just by watching someone else walk through it, right? And then, you know, when you take that knowledge and try to implement it, they're such a big gap. You're not going to hold all that information in your mind. You're going to have to walk through it to be able to learn it and walk through it multiple times, you know, even hundreds of times before you can start to gain an intuition for. The other thing too is in the same way that I ask Google questions, I also ask people questions. So, you know, if I have a challenge that I'm facing, I try and identify like, well, who would know about that? And let me just ask them. I also, you know, post online questions, right? Like, just really having that ability to say, hey, like, I'm not going to worry about what anybody thinks of me. I want to learn this thing. It'll be good for everyone if I learn it. So, I'm going to just go for it and ask that question and not feel intimidated to ask. Yeah, yeah. And that is, you know, when you work around, and really it's not just around people at your own work. If you have access, like everybody who's listening to this podcast can email me, for example, right now, you can email, you know, tons of people have their email just out there, but also Twitter, right? Stack Overflow. There's plenty of places, forums all over the internet where you can go and ask these kinds of questions, even, I don't know if you were developing at the time that this is popular, but before Slack, there was IRC. And I used to hang out in IRC rooms, the jQuery room, for example, Paul Irish used to be in there and he would answer questions. It was like this, you know, this kind of a towering figure now that feels a little bit disconnected from my reality. And in IRC and really on the internet in general, people are more accessible than we tend to believe. And it's easy to put people up on a pedestal, but I bet, Eleina, you would say, hey, you know what? There's an open line to Airbnb. And Airbnb is publishing information all the time and you're interested in people learning this stuff. So I would encourage people, and Eleina, you can speak to this too, to reach out to people that you don't know for helping this stuff. Sure, definitely. And I mean, going to meet-ups is also a great way to actually make that connection too. And then, you know, going forward, you have someone that you can bounce ideas off those. So unfortunately, Jonathan, I think we're going to have to take a rain check because I have to add to another meeting. I'm so sorry. Not a problem at all. Thank you so much for your time. I appreciate it. Thank you. And, you know, feel free to follow up with those questions at some point in the future. It's been really fun talking with you and I'd love to do it again sometime. Absolutely. Thanks, Eleina. Take care. Thanks so much for listening to today's episode of Developer Tea and a huge, huge thank you to Eleina for joining me on this episode. And for actually, for doing two separate meetings with me, two different times because of the original technical difficulties that we had, it was such an enjoyable discussion that I have with her and she's changing the way that Airbnb sees data and helping educate people about how to see data. It's such a good thing to have people like Eleina in the position that she's in. Thank you so much for listening to today's episode. If you found value in today's episode, if you enjoyed what Eleina had to say and enjoyed what I had to say, then I encourage you to subscribe in whatever podcasting app you use. We do have three episodes that come out a week. So it's very easy to get behind and then ultimately to totally stop listening to it at all. So if you think that these kinds of discussions and this kind of thinking and the challenges that we pose on this show are going to make you a better developer and a better person, then I encourage you to subscribe. You know, we're coming up on 2018 and I've been planning for next year and I think that this is going to be the best year yet. I really believe that. I think the content that we have planned and that we're planning for the upcoming year is going to be the most helpful content that we've ever published. So I'm really excited about it and I hope you are too. If you don't want to miss out on it, again, make sure you subscribe. Thank you so much for listening to today's episode. Thank you again to a few's for sponsoring today's episode. If you are building mobile applications and you haven't used anything but the default tools, and I recommend that you take a look at what Fuse has to offer. Head over to fuse slash plans and remember the code DTE can get you the professional plan for 70% off for the first 12 months. Thank you again for listening and until next time, enjoy your tea.E.