daliana liu_mixdown.mp3

Harpreet: [00:00:00] But everybody. Welcome to the Artists Data Science Podcast, the only self development podcast for data scientists. Our guest today is a data science leader with seven years of experience. She's also a content creator and podcast host. Hailing from Dalian, a coastal city in China and having lived in L.A. and Seattle before moving to San Francisco. She's currently surfing the waves of machine learning and AI at a W as her stellar career includes building the first time series forecasting model for advertising at Boingo Wireless and analyzing and helping launch over 100 a B test at Amazon. And if all that wasn't impressive enough already, she has a U.S. patent on experimentation that she holds with her colleagues. Currently, she's building machine learning models for customers from various industries to accelerate their business and products. But you might recognize her from LinkedIn, where she shares a data science, wisdom and career tips. Or you might recognize her as the host of the brand new stellar podcast called The Data Scientist Show. So please help me welcoming our guest today, podcast host extraordinaire, the incomparable Darlena Liu. Darlena, thank you so much for taking time out of your schedule to come onto the show today. I really appreciate having you here.

Daliana: [00:01:31] Yeah, thank you for inviting me.

Harpreet: [00:01:34] My absolute pleasure, man. I've been trying to get you on the show for. I don't know how long, man. I remember messing you back in mid 2020. Please come on to my show. Please come on my show.

Daliana: [00:01:43] I need to learn from you. You follow up?

Harpreet: [00:01:45] Yeah, that's what I do. Do you remember that one time I created a song for you? It was like to the to the tune of that song by llamo. I was like, Darlena, I remember, I think it was in the chat messages look it up. Yeah, but yeah dude, super excited to have [00:02:00] you here. I've been following you for quite some time. I'm a fan of all of you.

Daliana: [00:02:05] Yeah, also a fan of you.

Harpreet: [00:02:06] Off topic. You appreciate that. Let's learn a little bit more about you. I know I shouted at your home city in the introduction there, but. But talk to us about what it was like there growing up.

Daliana: [00:02:18] Yeah. So I grew up in a city called Dalian in China. It's in northeast China. It's a coastal city. There's a very obvious four seasons. So growing up we played with snow. In summer we go to the beach. So I always like living in a coastal city. That's why after I moved to the US, I lived in L.A., Seattle and now San Francisco. So growing up, my parents are really typical Asian parents. I like to read, but they always ask me to go out play. How fun. They don't want me to stare at their books too much. So growing up, I have the opportunity to explore my hobbies. I like to paint. I took some piano classes and really had a pretty happy and stress free childhood. And it's a coastal city. Of course we eat a lot of seafood, so yeah, always love to have some fish sushi. So that's what is it like growing up in China in DOT in it?

Harpreet: [00:03:23] Definitely the coastal vibes. Are you surfer at all? Did you say.

Daliana: [00:03:26] I only surfed once but I got a really cool picture snack so they got.

Harpreet: [00:03:32] That. And this is like the most boss move ever. I don't know if you've seen the movie like The Godfather and The Godfather. You know, he takes the last name Corleone, which is the name of the city that he grew up in. And then you took on the name of your city, seems like.

Daliana: [00:03:47] Talk to us about that. So yeah, basically the city is named after me. That's what I usually say. So yeah, I have a Chinese name, but it's very hard for people to pronounce and then I [00:04:00] still want to maintain identity from my roots. So I found Davina is actually a name, although it's probably a Spanish name, but it's a beautiful name. So I took that name and also it's easier for people to remember. That's how I got this name. Thanks for asking.

Harpreet: [00:04:19] Yeah, no, absolutely. Love that, man. It's such a like. It's like, the most awesome of everyday. I wish I could name myself Sacramento that wouldn't fly as well with me. So when you were in high school, like, what did you think your future would look like?

Daliana: [00:04:34] Yeah. In high school I went to high school in China, so it was very competitive, a little bit stressful. So actually I don't think I think about my future that much is more like you study all the time, you try to get to a good university. But one thing I always wanted to do is I like design, I like to write. So being a writer is one of the things I wanted to do, which is interesting. I'm doing a lot of writing today, either as a data science or as a content creator and as a designer. I didn't get to study design in college, but I got very excited when I started to learn data visualization. Oh, I can finally use some design muscle to decide what color I want to use, what type of design. And if you think about writing any data science analysis document, you have to use some design thinking to see how can you present the information to your audience? How can you make them understand what you're talking about? So it's interesting to see how what I wanted to do in college influenced what I'm doing now. But, you know, even data science is a new field, so there's no way I wanted to be a data scientist when I was in high school, and I think [00:06:00] I liked math, but I didn't think I was studying math in college or make a living out of it. So I would say I probably didn't see what I what I'm going to do today when I was in high school. So how do you.

Harpreet: [00:06:16] Oh, yeah. I mean, my road today is long and windy, dude. Like. Yeah, like, my journey started, like, you know, realistically, 2001 after I, like, messed around for a bunch of years, bounced around and did bad stuff and then got back on track. But for me, like, like you said, data science never was really the thing back then, right? So I graduated high school in 2001, and after I graduated high school, I just went on like a very bad path for several years and then finally got my act together. By this time I was like 21 and it was like 24 when I decided to go back to school. Then even then when I went back to school, like I just get degrees. That was my mentality, right? Yeah. And then ended up going to like grad school and all that stuff. So my vision for myself was I was going to be an actuary because I was like, Okay, well, I like math. I'm kind of good at it. What's the the job that I can get that will get me paid with this particular interest and skills that I have. So back then 2011, it was actuarial sciences, say a full ten years ago. So then I went to grad school and took a bunch of exams and then started working in actuarial sciences and realized that it was actually predictive analytics and predictive modeling that I like best. And then other stuff happened and got drilled off that path and now I'm back. So yeah long and winding road to get to get to it to to this point. But yeah I always loved I always loved math. So I guess how did you kind of gravitate towards math having this interest that seemed a bit like honestly, like artsy kind of interest. Like did you see any kind of art in mathematics? Is that what kind of drew you to it? Like what? What, what was that pool?

Daliana: [00:07:55] Yeah. So it wasn't the art of it. It's more like I didn't [00:08:00] know what I wanted to study. I wanted to study design, but I also wasn't so sure about that. My parents want me to study math. They think if I later on want to do something else, math can lay a good foundation. So I just listen to them. But when I started to study math in college, it was hard. I didn't like that. It was very stressful and I even resented my parents for suggest me suggesting me to study math. So then I didn't know what I wanted to do to find a job. That's when I started to think, okay, maybe I should study statistics. By the way. Also, a lot of my classmates went to study, actually, and I think statistics more tangible at that time for me because to deal with a lot of uncertainty. So I got excited about using that to solve real world problems. And then I think at that time I started to enjoy what I was doing more. So that's kind of my journey. I and also like a side story is the reason I have to pick math as a major was because in China we have this college entrance exam, you do the exam and then you take measure. I did a terrible in math so that dragged my entire score down and I didn't have a lot of options.

Daliana: [00:09:25] So it's kind of ironic. I didn't do a math and ended up studying math and then it's very interesting. So I think when you're in school, when you started your journey, you always want to do more advanced fancy stuff. When I started working, I always want to learn more machine learning, deep learning, all that. But now they started to go through some old statistics or math textbooks. I started to appreciate the art of statistics and math, and I wish I learned more when I was in school. Because at end of the day, if you really [00:10:00] want to give people advice, want to be creative in your solution, you have to get good at the foundation and it really takes multiple time to really understand something. So now I think I can finally understand the art of math or stats. It would be weird. I used to dread doing homework, proving the theories, but now sometimes I pick up a book. Maybe because I already understand most of it. When I look at that, it's kind of like entertainment. Maybe you read some statistics, book review, some statistical testings after work over the weekend. So it really changed the mindset.

Harpreet: [00:10:40] They actually love that. I like to revisit the basics as well, but I like to do it in as fun a way as possible. So I've stumbled upon these books that are like the manga guides and like the Cartoon Guide series. So there's like the Cartoon Guide to Calculus. Cartoon Guide to like, really? Oh, it's so good. Yeah, they're amazing. Just like the Cartoon Guide to Calculus, Cartoon Guide and computer science and Manga Guide, all these things. But I'm a huge fan of that. And like, like I love, love, love, mathematics. But I feel like as data scientists, we don't even get to explore like the most interesting and weird and fun parts of mathematics, like geometry. I love geometry. But then there's, you know, when we think of geometry, mostly things like flat geometry, but then there's like topology and geometry on like different surfaces and then there's like prime numbers and number theory and Riemann hypothesis and all this crazy stuff out there that I feel like we should probably do a little bit of research on. And then even like the philosophical aspects of it, like, you know, incompleteness theorem and things like that, they're so fascinating. Yeah, I love math.

Daliana: [00:11:42] Yeah.

Harpreet: [00:11:43] You mentioned the writing a couple of times. And you know, actually, I've noticed this you've made this a point in some of your most recent content is the importance of of writing. You mentioned that as not only as a content creator, but as a data scientist, part of your day to day includes [00:12:00] a lot of writing. So talk to us about that. Like, how does writing fit into your day to day? How is how is writing been an important element of success in your career?

Daliana: [00:12:12] Yeah, that's a good question. If you think about data science work, a big part of it besides getting data, a big part of it is to translate the business problem to a data science problem. And I think being able to write it down in complete sentences instead of bullet points really help us to think about things clearly. Right? What is the standard? What is the measure of success? And why are we doing this? Sometimes, maybe halfway. When I'm writing something, I really owe the whole approach. I'm thinking about this is probably wrong. So I think writing is the way to help us think clearly. So that's very helpful when you are approaching a problem and now when you completed a project, you want to present it to your stakeholder. You also want to write about your conclusion and maybe you want to use different type of languages or different type of layout. When you send this report to business owners or science peers, I think is also important because a lot of people don't have the background of the problem you're solving. And if you don't talk about it in a way that easier for people to understand, they might be confused. They don't know how to give you feedback. So I think that's another important part. And and also a lot of times you want to write down some best practices. So for your team to review this document, transferring this knowledge so you don't have to always teach people how to do things. And during the process, when you communicate with stakeholders, you send emails [00:14:00] or slack messages.

Daliana: [00:14:02] This all is writing. So the core of is how can you make something complex, simple, and how can you use the other person's language to speak to them? And another level of it, to be a good writer, you have to have a lot of empathy. So basically what you're doing is make the other person's understanding easier. So a lot of times we think if you are good writer, you write a lot. There's some mistakes I made. You present all your findings in the long ass email and send it to people, but it actually make it very hard for people to understand. So they're saying that what is it? It's I don't have time to write a short email. Writing the short email is hard. You have to think, what's the most important thing you want to talk about in the beginning? How what is your call to action? What type of response you want to get from the other people? And a lot of times, signs, findings are nuanced. How are you? How are you going to present it without mislead people? So if you're able to summarize your entire research in say 200 words and send to your stakeholders and they can understand it, they can give you feedback or even, you know, on board with the next steps, then that's very successful. That helps you to get resources as a data scientist and also help you build influence for for your team. And basically, people will actually use your solution and that's very powerful.

Harpreet: [00:15:41] Absolutely love that that that's so many good points in there, man. Especially the part writing is thinking like writing. Help to clarify your thoughts. Yeah. That quote I forgot who was, but it was like if I had more time, I would have written a shorter letter.

Daliana: [00:15:56] Yeah.

Harpreet: [00:15:57] How do you. How do you suggest people get better at writing? [00:16:00] Is it just like going through, like a business writing class, like one of those free business writing classes? How did you get developed that skill?

Daliana: [00:16:08] Yeah. So I think first I'm interested in writing, so I pay attention. When I write articles. You see how other people construct their either science blogs or maybe some newspaper articles. So pay attention to the type of writing you find that speaks to you. Everybody have different style. There is no universal style. So if you really like a writer, read what this writer is writing about and then learn about their their style. And another blog post I found a very helpful was, Do you know this writer? His name is Scott Adams.

Harpreet: [00:16:49] Yeah, yeah, absolutely. Yeah. The day you became a better.

Daliana: [00:16:52] Writer, the day you've become better at is so short and it's very powerful. I think maybe we can just share what we can remember with the audience. So basically remove. Principle number one remove useless words instead of saying something. It's very good. Just say it's good. And what are some other principal? I think this is probably just the most important thing. And when when you and also don't write long sentences, it's very hard for people to combine all those pieces, try to see if you can break it down. So it's very easy to write long sentences but have the awareness is a tool and see if you can break it down into a few sentences. And also don't use the passive verb write to say who I. I want to do something instead of saying like something is being done. And also if you talk about creative writing. If you sometimes maybe you want to use a verb that has more emotion, that's more exciting. So does some word picking that might not be useful for scientific writing, but I do [00:18:00] think it's helpful to make the sentences shorter. A lot of times when I write on LinkedIn right now LinkedIn posts because I love that LinkedIn has that word limit, right? So if you go past the 1300 word, you cannot post it. So I have to trim it. So I spend a lot of time trimming my sentences and then I realize, Oh, instead of writing a full article, maybe I just need three sentences. People will get what I'm trying to say. So whenever you're writing itself, academic is shorter, can make it more clear. And I think those are some very important things you do that you're already better than a lot of people. And another thing is to get feedback similar to doing your data science project. You don't have to get feedback when you feel you're ready, you're finished because it would never be perfect. Just once you have a first draft or even just a beginning or outline, ask people what do they think? I think having that feedback and iterate is also a good way to write better articles.

Harpreet: [00:19:04] Absolutely love that the baby became a better writer. Go look that up. That will make you instantaneously, like literally a better writer. Within a few minutes, it'll take you part of the way. There is still a lot of practice, but that book really got me into writing a business writing after I met that book. But that blog post after that blog posts have picked up like Business Writing for Dummies went through that entire thing and a couple of like LinkedIn courses on business writing and it just they help so much. Yeah. Excellent, excellent recommendation. I love that. Have you read any of these books?

Daliana: [00:19:33] Sometimes, yeah.

Harpreet: [00:19:34] I feel like you're in my head sometimes.

Daliana: [00:19:36] Yeah. I read his book. What is it? The one. Golf strawberries.

Harpreet: [00:19:41] Oh, yeah.

Daliana: [00:19:43] Oh, that was so good. You talk about.

Harpreet: [00:19:44] Crazy.

Daliana: [00:19:45] Probability. What if probability is God's language? I'm like, Woo!

Harpreet: [00:19:49] Yeah, that book is insane. I love it. It's short too. It's like, I'm looking at it right now.

Daliana: [00:19:54] Yeah, I highly recommend you love math probability. And if you think about [00:20:00] live, do we live in a simulation? Have you ever thought about that kind of stuff? Highly, highly recommend.

Harpreet: [00:20:06] Yeah. Yeah. You do a simulation. Diana, what do you think?

Daliana: [00:20:09] I think we mind. It's like there's no way to prove we're not living in a simulation.

Harpreet: [00:20:15] Yeah, I was watching some YouTube videos about, like, kind of just, like, the nature of consciousness and what it means. And, like, technically, all of us are living in simulations because we're not actually seeing a reality very objectively. It's being filtered through our perception and our experiences, and it's just crazy to think about that stuff. We can go off the deep end for that man, but I.

Daliana: [00:20:38] Feel like that's another one. Our conversation.

Harpreet: [00:20:40] Exactly, man. We got one of these days, we'll just go off the deep end. We call it the Harp and go off. Yeah. So let's talk more about so we talked about writing, but what are some other critical elements to success for someone's career as a data scientist that don't get taught in school?

Daliana: [00:20:56] Mm hmm. Yeah, I talk a lot about that. I just feel there's a big gap between school and real world when you're working in industry. I think one thing is in school you always have a perfect answer. There is a right and wrong. You have a grade, but in reality, a lot of the solution you developed, maybe there are other way to solve it and you solve it in this way doesn't mean it's wrong. So I think have the mentality that there might be different different ways to approach a problem. Help us be more confident when we deliver a solution. If your solution meets the business goal, solve the customer's problem, then it's a good solution. And people say there is no perfect model. There is a model that solves the problem. And also I talked about before in school, you finish a course, right? You submit your homework. But in reality, you need to constantly get feedback from your manager, from your peer, from the stakeholder. [00:22:00] I think this mentality to build something crappy is not good, but I need to show it to people and then I can iterate. We don't learn this a lot in school and even some school project is usually you spend a week, you just kind of crash it and you don't really learn how to get feedback and also how to form your hypothesis, right? There is sometimes no clear way to translate a business problem to a data science problem. Then you need to learn how to ask stakeholders, Hey, how do you measure success? You say you want something good, what is good, right? What is something that keep you up at night? So I need to think about basically I make this meme like data science that like Sherlock Holmes, you need to figure out all the pieces of information.

Daliana: [00:22:53] You're like investigator, but instead of having any weapons and stuff, your superpower is data. So we don't learn those type of things in school and also in school. To solve a problem, we might focus on accuracy and in reality what we're of course, we still want to have a good model, but what we think about is impact. Sometimes maybe you don't need a complicated model, but you want to develop something that really save the business money, save people time, and I think shift that mindset from policy to business. Impact is is also important, but I don't think the school is completely run about it. I think we still need to have a good foundation in a relatively simple, pure environment to learn data science, to learn math, and now overcome the young minds so people can just be curious and explore. But I would really wish, maybe say for data science major students, maybe third or last year, they can have some sort of project [00:24:00] and then to explore, make mistake. And get feedback. Collaborate. Write a blog post published on GitHub or Medium to really understand what does the end to end data science project look like? But don't be scared that if you feel you have no clue, everybody starts from where you are right now. Everybody experience the same struggle. So just don't be afraid to ask questions and think about the business problem. On the too high level, I think you will learn very fast.

Harpreet: [00:24:34] So how did you learn these skills? So these are skills obviously they can't be taught. So how did you pick up on these skills?

Daliana: [00:24:43] Right. I don't think I can just learn those skills. It's more like I have made mistakes. For example, sometimes I procrastinate on the project because I don't know how to solve it. And also I feel, Oh, what if other people think I'm stupid if I go ask questions? So I never ask questions. And in the end I present this to my coworker, to my manager, and they give me some feedback. Then I realize I should have wasted a lot of time if I have asked those questions earlier. I can produce something maybe two times better and nobody really told me I'm stupid for thinking about those problems. And then I'm very fortunate to have worked with managers who are very honest about what they know. They're all experts in their domain, but they're telling me, Hey, you're solving a new problem. I don't know. I don't know the answer. And but we can we can brainstorm. We can find other experts to see how we can approach this. So in data science, probably in a lot of tech domains, because things move really fast and we have no problem every day, we have to learn to be comfortable with this other skill.

Daliana: [00:25:57] Why are you building it? So in the beginning, I feel very [00:26:00] insecure about not knowing how to solve a problem. I have the illusion that I know all the regression machine learning testing, but I don't know how to fit in and the problem. And then in reality. There. There is isn't really this magical map that, oh, for this type of problem, you do this, that problem, you do that. Everything is different. Sometimes you have to take a hybrid method to do things and you just need to be creative. Sometimes you need to trust our intuition, but don't trust it too much. Like I said, I got feedback and you have to work on the project. You have to ask your manager to give you a project that make you feel a little bit uncomfortable to exceed your your level so you can grow and learn. And you have to do this by making mistakes, but still withing the level that it's not going to ruin the business. And through those mistakes, you learn. You grow and you have you got your lessons.

Harpreet: [00:27:08] That's such good advice. Thank you so much, Diana. I'm wondering if a little bit in there is that feeling of imposter syndrome, a feeling of not wanting to ask a question because you don't want to be perceived as not knowing something like, oh, you're supposed to be a data scientist. Don't you know how to do this? Do you notice this happening a lot with with data scientist of any career level?

Daliana: [00:27:35] Yeah, this is imposter syndrome. I think it happens for everybody. As long as you learning new things, you're taking on new challenges. You always feel you don't know this type of thing and you're scared. Other people might think you're not good enough. So again, you can get feedback, right? And maybe when you ask a lot [00:28:00] of questions in meetings, you realize, okay, I ask the question, but basically everybody else is scared of asking. Probably you doing everyone a big favor by asking the question and really, what is the risk of asking a stupid question, right? And maybe you look stupid, but if you don't get an answer of it, you also look stupid for longer time until you figure this out. So I don't know what's the solution for that. I think you just need to be comfortable to look stupid from time to time, right? Ask that question. And also, I think there's something an organization can do know. There are some people I work with there. Sometimes they just throw in a lot of information at me and they don't have bad intention. But I didn't realize, okay, I actually have a gap.

Daliana: [00:28:51] I don't know everything. You can just give me a script and I would know what's going on in your code. And then there are some other people I have worked with that have the awareness there. Hey, by the way, I know it takes a lot of time to figure this out. Don't be afraid to ask me questions. So I have learned that that's very helpful, that make me feel comfortable to ask questions and look stupid. And when I start to mentor people, guide people, I've learned to say things like that, to try to understand where they are when I start to work with them. So if you know so as a data scientist, we collaborate with other people like product managers or software engineers. So if you don't understand your work, don't be afraid to to ask them, but also build a relationship with them and make them willing to teach you and vice versa. When a product manager or software engineer asked you a question, remember the time that you feel insecure and be more patient, have more empathy for people.

Harpreet: [00:29:58] Yeah, absolutely. Absolutely. Love that. I think the biggest [00:30:00] shift for me my career came when I just realized that actually nobody really expects me to know everything. And not only that, if I was to ask a question, that would be stupid. I might be thinking about it days later, but I'm pretty sure the other person just forgot about it. Yeah, right. Like they just. It was just whatever to them. Like, they're too busy thinking about themselves. Yeah. So there's no downside. Yeah. Ask the questions.

Daliana: [00:30:23] Right. And I also want to add to that, if you spot somebody who seems like they know everything, they have an answer for everything. That's probably a fraud.

Harpreet: [00:30:35] State of I don't know is my favorite response to it. Everything that becomes really hard, man, because like people come to me for like these office hours that I do and they just ask them questions. And like my initial response, I was like, I don't know, man, you got to tell me more. You know, this is, it's, it's yeah. Anyways, let's talk about how your day to day work as a data scientist. How is this different from what you expected it to be when you were an aspiring data scientist?

Daliana: [00:31:04] Yeah, so when I was looking for a job, I thought data science or data analyst probably just stare at a screen. There are beautiful charts, there are some some tables and I just sit there, analyze it. I didn't know what I mean by analyzing data. I just feel like, Oh, you come out with some cool ideas and then you pitch to the business, right? I have the image of I'm wearing some fancy suit and go present, convince people every day. So that's what what I was thinking about the data scientist day to day look like. But actually what you do is it really depends on what project cycle you're on. For example, in the beginning you don't even get to look at the data or you don't know what data, whether they have data. You spend a lot of time in meetings to understand the business, understand the constraints, [00:32:00] what type of cause they're comfortable with, what type of investment they have, are they going to use it? Do they have leadership buying? You're asking a lot of questions and then you try to distill those questions into your mental model to decide, okay, what's going to be my next step? Is this project worth my time? And if it is, what type of resources do I need to make it successful? Right? So you're doing a lot of things like that.

Daliana: [00:32:27] You ask a lot of questions and once you finish that phase, you start to gather data. So at that time you try to get a lot of permission. So you're still probably going to be in a lot of meetings with product managers, maybe data engineers, software engineers to gather the data. And then you slowly move to the data exploration and modeling part, and then you start to write some code. But at this time, I still talk to stakeholders to get feedback, to ask them, Hey, is this what you want? Or, Oh, you want 100% accuracy. My model cannot do this. What type of compromise do we need to make? And at this time you do a lot of iteration. You might create ten models and you collaborate with other people, maybe more, maybe your science community. And in the end, when you finish something, you start to communicate with your stakeholder more. And you if this model is going into production or some type of solution, go into production. You talk to some other downstream teams that could be engineers, product managers again.

Daliana: [00:33:39] So as a data centers, we always have a lot of meetings in communication and sometimes it's hard to really have the time to build your model, analyze the data, and a lot of times the model is already built, you just need to tune it. So sometimes as a data scientist, [00:34:00] I feel I'm doing a lot of talking and also some engineering work. And in some projects actually you don't get to do a lot of really math stats work. So that's common. And in some other cases, maybe you are you do need to understand very deeply in the math stats and the models. So I would say really it really depends. I think it's important to have the right expectation. Also understand data science job is to solve the business problem and not get easily frustrated. If you don't get to play with the math or some other part you're curious about, I think it's important for us to have the balance. If you're not really learning or using a lot of things that you find exciting, maybe set aside some time for yourself, say on a Friday afternoon or after work to satisfy your curiosity. So then you don't get burnout from this type of work.

Harpreet: [00:35:02] My next question was going to be, what are five things only experienced data scientists? No, but I think you've listed off a bunch of them. There's a lot of good gold nuggets there. Definitely pause and go back and rewind that because I have some good stuff there. So let's skip that question about five things only experience data scientists know because he gave us like 12 of them just now. What what do most data scientists, I guess, do wrong when it comes to their career development?

Daliana: [00:35:29] I think I wouldn't say I know the answer to the the the one thing I would just share a few mistake I have made, I think in my career development, I think sometimes I'm afraid to ask business questions, to challenge their assumptions. Maybe someone tell you, hey, I want to do something, and then you're like, okay, you put your head down, tried to develop a solution, and then halfway [00:36:00] through you realize maybe their leadership changed. They have different goals. What you're working on is not relevant anymore. I think it's important to always have this type of alignment with your manager and know that your people say, Oh, stakeholder alignment. You do that in the beginning of the project. Also, you need to do that throughout the project to make sure people still want the thing you are working on and is still. Then I think to have a successful career you need to bring impact to the business, but also you need to be interested in the work you do to so sometimes maybe aware, just really want to use the most advanced model we just learned and the way for God, oh, this is really not the business need.

Daliana: [00:36:51] So I think to balance the impact and also your passion is very important. So sometimes we'll forgot one or another. And also I think data scientist sometimes is not forget to. So one mistake I made is to forget to measure the baseline. So if you're solving a problem kind of ready been solved the menu solution or you're taking over someone else's problem, they want you to improve it. It's very important to get a baseline. Okay, what is the current metric? Not just the accuracy, precision, recall or whatever, also the current impact. So if you don't know how to compare your solution to the previous solution, then how do people evaluate your contribution? So I think knowing this baseline of your work is another thing that's very important. And I think a lot of time they design is forget to gather this data point that's both important for you to measure the impact of a project and also from a career development perspective, you need that data point [00:38:00] when you write a promotion document or do interview.

Harpreet: [00:38:06] At that point about baseline is a super, super important because I think you might have made a post about this if I recall it was talking about like the question that what is the best model or how do I pick the right model? Well, there's this concept of the multiplicity of good models, like any number of models can perform hypothetically well on data. It's just a matter of finding the baseline and then additively improving on that. So that's definitely an excellent point. So I guess where would you draw the line between, let's say, a data analyst and a data scientist? Is it like can you point to one skill and be like, Oh, right there, that's it, that's it. If only you knew this one thing, you'd be a data scientist. Does it work like that? What are your thoughts on that?

Daliana: [00:38:51] Yeah, that's a tricky question. I think intuitively I would say if you are using a lot of scientific method, your data scientist or I think the stereotype for data science and data analyst, data analysts just measure things, create a business report. I observe this trend. Business is doing well. And then data data scientist comes and asks the reason, okay, is this trend real? So data analyst discover a pattern and data assigned to would validate whether this pattern is real or not. So then because to evaluate whether it's real now, maybe you need to do some experimentation. You need to know AB testing, you need to know causal inference. You need to develop the model to see what's really going on there. But also observed a lot of data analysts these days can do those type of things as well. And now also I see some data scientists, they don't do a lot of modeling their work, focusing on reporting, building metrics, working closely with product managers. And that's also very important. So I would [00:40:00] say the line is blurry right now and there's really no good about it depends on what you are interested in. So don't focus on the title. I think both the industry and the recruiter still figuring things out. When you're looking for a job, look at the required skills, focus on the skill set, and also look how the tools they're using are the most important to they say is say, for example, TensorFlow. Then that's probably like machine learning, deep learning job. And if they're using a lot of SQL Excel, the data analyst or BI engineer type of job.

Harpreet: [00:40:37] Absolutely love that that statement he made. That's kind of how I think about it. Like an analyst analyzes that, a scientist discovers. So.

Daliana: [00:40:47] Oh, yeah, you put a beautiful way like that.

Harpreet: [00:40:50] You know, I think an analyst just kind of uses data to understand the behavior, whereas the scientist will use experiments to understand some system, right?

Daliana: [00:40:59] So one observes another one kind of, I don't know, discovers or challenges or innovates.

Harpreet: [00:41:07] Something I've been wrestling with kind of philosophically. You know, we're talking about you talked a little bit about scientific methods like what is the science in data science? I've been really trying to wrestle with this like, you know, just deeply philosophical question, I guess is like, are we actually doing science? And I mean, I guess from a viewpoint of something is scientific. If we can make a hypothesis and then that hypothesis is able to be disproved, then we can say that this is proper science. And I think to an extent, yes, I mean, definitely to a 100% large extent, we are able to do that. But I'm positive that I'd love to get your thoughts on this.

Daliana: [00:41:43] Yeah, I definitely had the same thoughts. Say, if I'm working in a startup, all I need to do is to. Build a data pipeline, measure the data. And does it mean I don't deserve this data scientist title? [00:42:00] I think you can still be a data scientist. Data scientist in some company I even see. I even have seen data science engineer this type of role. I think the word science in this era can be pretty flexible. I think the science just means all the elements that's related to utilize data to make better decisions. So you either use data to make disease decisions or you use the data to simplified some processes. So as long as you are doing something like that, I think it's data science. So basically everyone who's playing with data is data science. But I think that data science compared to just analytics, I think there's definitely a element of, like you mentioned, how the hypothesis and test the hypothesis is not just, okay, this is a data I collect and then you immediately have the story. You would do some investigation using statistical machine learning those type of tools to date the the hypothesis and not to fall in the trap that if I see something interesting, there is something going on. Sometimes it's just it's random. I think with that critical mindset is what separates data scientists from other data roles.

Harpreet: [00:43:26] Love that. And there's a point in there that reminded me of this other post that you made a while ago, and I'm 100% on board with this. It's just call yourself a data scientist, like it's a permissionless field, really. Anyone can download Python, anyone can find data, you can start coding, you can set up a SQL Server locally on your desktop. You don't need anyone's permission to do these things. And I was just surprised that like some of the the amount of pushback and people commenting on this right. In a negative manner, like I'm 100% on board with the I've been saying this to my mentees for a while. Like you [00:44:00] just, you know, if you are practicing data science by doing data science work, even though you're not getting paid for it, I can call yourself a data scientist, like no big deal. Hmm. What are your thoughts on why people are giving you so much pushback around that particular thing?

Daliana: [00:44:15] Yeah. So before I answer the question, I wanted to say, where did I got the inspiration of the post? I saw someone, probably a designer, saying, if you open Photoshop, you're drawing something, you create some simple design. You are designer, right? And then I was like, Wow, if someone just created one design using Photoshop, I wouldn't question whether they're designer or not. Right. But why we have this type of kind of gatekeeper thing for data scientists? I mean, yes, data science is hard learning. All the math of statistics is difficult. And like I mentioned, in order to adopt that critical thinking, you have to go through some training. But a lot of people become data scientists coming from. Some engineering major or I did a podcast with someone who studied history in college and then they got into the design through some gradual learning. Maybe they initially got interested in understanding the business, so they play with some SQL and Excel, and gradually maybe they took some online courses and some of them feel okay. Now I also want to get a holistic view of the data science field. I do a master program then, and it's really hard to say at which point they become a data scientist, right? It's probably not a point that I'm hired by a company. I think they probably already have the skill set before. So I think I got a pushback because a lot of people just read it, read the words literally.

Daliana: [00:45:51] They think, I'm lowering the standard. I think it's not. It's more about if you want to become something, you can have that standard for yourself, [00:46:00] right? If you're you think your data scientist, what do you need to do to meet that standard? And also we talk about imposter syndrome, right? You don't need to feel you're not enough just because you're learning. I'm still learning. If you are asking me some interview questions without preparation, I probably can't answer as well as you are. So a data scientist, someone who solve a problem and draw business value and then apply adopt this critical thinking. And to constantly question is this pattern true? Is there something wrong? So if you have that mentality, you are a data scientist and you don't need to let other people to discover, as you and I have worked with so many great data scientists, just a lot of them didn't have master's degree and or did a bootcamp. And also I'm not against grad school. If you feel grad school helps you to learn deeper, I encourage you to do so if you have the budget and time. But if you feel you have learned enough in work or through books you read, and if that's enough, you don't have to just feel not enough just because you don't have that degree.

Harpreet: [00:47:13] Yeah. I definitely learned a lot more outside of grad school than in grad school. Like, I left grad school with not very many skills, which is sad to say. And you know, data science is a career you're perpetually learning. You have to be comfortable with that. You have to be comfortable always learning. There's never an end point. Right. It's you know, if you decided for this career, you are at the beginning of infinity and it's a beautiful, beautiful infinity. And there's like two classes of career. Like, I remember just reading some of the comments. I'm like, okay, this guy's talking about, right? Like there's two classes of careers, really. There's careers where you legally have to go through a program to be called that. Right? Like nobody nobody can fucking take a a mook and start doing brain surgery like that doesn't happen. Right. But [00:48:00] you can't take a series of online courses and start practicing dentistry. You can't even do that and become an accountant or become an actuary. Right. There's certain fields where there's these hurdles and these keepers of the gate, but in tech this doesn't exist. Nanotech is permissionless, it is permissionless to get into this there. Definitely thanks for sharing your experience that sorry your viewpoints on that 100% on board with you with that kind of mentality and that kind of thinking. Can I do that for myself? And like I was a biostatistician for five years before becoming a data scientist, but in order to boost my ranking in LinkedIn searches, I just changed my headline to say Data Scientist. Right, and more opportunities are coming my way. It's not like I couldn't do the job. I had all the skill sets. I didn't need somebody to say, Oh, you've never been a data scientist before, right? Yeah.

Daliana: [00:48:47] Yeah, that's a great point. Can I just add to that and you should change your LinkedIn title to the job role you're searching for. And a lot of people put things like, Oh, looking for a full time job in your title. That's not helpful. Where Recruiter are searching for people their your title is very important they're looking for the keyword so instead of saying looking for a full time Rhodes that put skillset or your data science emotional learning engineer and I love that you shared a story you change your title right because people are looking for relevancy so the I got into Amazon because I have a LinkedIn profile though it's very simple. I probably have, I don't know, 200 connections at that time, but my job title at that time was a business intelligence data analyst, and the recruiter was looking for a business intelligence engineer candidate. So I guess there's overlap between business intelligence, but I think it was lucky they reached out to me because in some other company, maybe it was just called a statistician or data analyst. So if you are looking for a business intelligence role and you of course, if you believe you have those skills, [00:50:00] be intentional. Your current business should be business intelligence. If you have business intelligence, appearance or same for data scientists or machine learning machine learning engineer, maybe your current title is focusing on, say, risk of fraud or research. I have a friend that his business title is like Fraud Engineer. I'm like, what it's supposed to mean, right? Yeah, I get it. It's your title in your company, but it's not helpful for recruiters to discover you and your your major headline doesn't have to be your company's business title. You can stay in a job description in your resume and stuff. But the most important highlight, if you are looking for a machine, an engineer will put machine in engineer there. Don't put some. You know the things that recruiter won't understand.

Harpreet: [00:50:52] Yeah, absolutely. Like for me, like right now, like first thing on my headline on LinkedIn that says at the time of this recording, it probably stay there, says Philosopher. I didn't study philosophy in school. I didn't get a PhD in philosophy, but I read a ton of philosophy and I practice a lot of philosophy like I do a lot of philosophizing. And that's how I view myself. Like I view myself first and foremost as a philosopher above everything else. So yeah, I'm going to reflect that in my headline because that is how I view myself. It's a self image.

Daliana: [00:51:23] Yeah, exactly. And also that regime forces you to think more about philosophy because I believe you're not a fraud. I'm not a fraud. When you put it there, you know, you have to think a lot about philosophy. You read books, and then they also motivate you to really live up to the title. Right. So, yeah, actually, when when I just met you on LinkedIn, who is this guy? You know, philosophy. Philosophy. But now when I get to know you, I know is actually authentic to you. And now. So that's what makes you unique, right? I don't see anyone in data science if you put their [00:52:00] title like that. I you know, I'm not I think if you identify something, make you special, but own it.

Harpreet: [00:52:07] It's just one of the topic of Lincoln. And just kind of speaking from my own frame of reference, man, I feel like it can get incredibly tough to find a voice for yourself on LinkedIn and just do content creation, especially when you want to say something that hasn't already been said. How do you try to ensure that you're providing as fresh a perspective as possible with the content that you create?

Daliana: [00:52:32] Yeah. So actually it's funny when I'm creating content, I'm not thinking about if that's fresh. It's just I started because a lot of people asked me, how did I get into data science? So I think maybe it's easier for me to just share my experience. And then I start to realize there hasn't been a lot of people providing good advice and getting into data science. How do you grow your career? And maybe sometimes I do know some of my opinion is fresh because it's some idea stuck in my head for a long time and I'm really grateful that I have this platform that I can share my thoughts. Otherwise it's just going to explode in my head. And you know, when you go out with your friends, you can just say, hey, I have the thoughts about data science career. I'm going to tell you. So I would say when I run on LinkedIn first, I write for myself, is, is I want to express myself. And then I know it also benefits other people.

Harpreet: [00:53:33] So what are your thoughts on what it means to be to be a data science influencer?

Daliana: [00:53:39] Yeah. So I guess some people call me that, but I also feel is like a funny word. Like to be an influencer. I think it's more like I. Maybe I'm just a writer. I'm sharing my journey. I really love sharing the mistakes I made. I think, especially [00:54:00] when people say, Oh, you know, she's a senior data scientist working at Amazon and even she made those type of mistakes. And people will feel maybe I can also get their data science role is not intimidating. So I really try not to take. Data science influencer. Quote unquote, too seriously. I think I just kind of want to have fun with it. But if there comes with any responsibility, I think I want to create an image that people kind of say being a data scientist doesn't mean we were just thinking about math and coding all the time. We love nerdy jokes. We are philosophers. You know, we we have a life, we are humans and we make mistakes. So I kind of maybe more like wants to humanize data scientists. And that's actually a motivation of why I want to build a data scientist, show the podcast to give people a 30, 60 degree of how they become who they are today. Of course, what are the cool things they're working on, but what's their career journey? What are some mistakes they made? Who they are when they're not doing data science?

Harpreet: [00:55:06] Absolutely have that. And if you guys want to, Alyona, we'll talk nerdy to you. Her newsletter, I just read them. I think today was one that came out great newsletter. Definitely sign up for that. I'll leave a link for it in the show notes. We'll be sure to link to that and your website. But let's get let's get into your podcast though the data scientist show. Talk to us about that. Like how did that idea come into your mind that you want to start a podcast?

Daliana: [00:55:27] Yeah, I would say maybe you're one of the inspirations. You're doing a great job for the artists of data science. I think you tried to be the I think you're already the Tim Ferriss of Data Scientist podcast. You and I really love that you don't just talk about data science. You have a hybrid of content, you interview data science, you also interview someone who does something not related to data science, but interpersonal development. And guess what? Everybody needs personal development, right? And for me, I in the beginning, I always enjoy [00:56:00] having conversation with people. Like I'm really enjoying our conversation right now in the beginning, to be honest, before I come here, I'm a little bit nervous. I think maybe this is like the first or second podcast I did because I'm pretty selective in going on the podcast. Yeah, but I accepted the invitation because I already know you. I know we're going to have a great conversation. So I think, why don't I just do something I really enjoy and I know other people will learn from the conversations. So it's a win win because sometimes when I talk to my data scientist friends, I would be like, Damn, we have really good points. I wish we can record it that right sometimes I right write it down and share that on LinkedIn. But I think the audio format is pretty cool because if I imagine some of my followers going on a walk and listen to me talk, I just think it's it's something really cool and I want it to be easy for people to listen to so they can learn while enjoying this conversation. So that's why I also call it a show. It's not like a tutorial or something.

Harpreet: [00:57:10] I love that the format, like you're interviewing people live and direct, like I absolutely love that. That's such a that's so key. That's such a crucial element. I think with the video element of the podcast, the way you do that is that's absolutely amazing. So yeah.

Daliana: [00:57:25] And also sorry, I don't I don't know if you noticed this. They also give you an excuse to have interesting conversation with people you normally don't talk to, right?

Harpreet: [00:57:34] Yeah, absolutely. 100% like nobody else would want to talk to me.

Daliana: [00:57:40] No, I don't think so. For example, I met this great professor, Conrad Courtin. He's a professor in neuroscience and a mutual friend introduced us. Without this podcast, there's no way I can talk to him and learn how he decode the brain using AI. And it was his unconventional career. [00:58:00]

Harpreet: [00:58:01] Obviously I've got some great people on the show. Next thing was on there. You've had the. Was it Jonathan or Javier from one Sultan?

Daliana: [00:58:09] Yeah, I had early Jerry Lee.

Harpreet: [00:58:12] Yeah. And then there was the one that's on my playlist that's ready to watch was the episode about how to use data science combined with blockchain because.

Daliana: [00:58:20] Oh, yes.

Harpreet: [00:58:21] That's super interesting. What was the guy's name? Shut that episode up.

Daliana: [00:58:24] Greg Ossuary. He's the CEO of Cash Network.

Harpreet: [00:58:28] Yeah, yeah. I definitely tuned in to that podcast. I guess, like, who should listen to this podcast? Is it like, do you have to be, like, experienced in the game to listen to it? Or is this, you know, a broad spectrum of of data scientists like.

Daliana: [00:58:41] Yeah, I called it a data scientist show, but it's actually you can I interviewed Zach Wilson. He's a data engineer. And we'll talk about data engineer. Data science in general. If you're a machine learning engineer, your data analyst, you're interested in data visualization. So basically, anything related to data or career, I recommend you to listen to the show and just enjoy and learn. And also, if you're a product manager or your software engineer, you work with Edison. So I think it's also great to peek into the data on his mind and to see where we're thinking about. Or maybe you're a student, you want to get in the field and see what do people work on on a day to day level? Was their struggle, was their journey. I think sharing those journey is very important for the community because I can post on LinkedIn every day, but that's just my story. And imagine how powerful it is when you learn 100 people's journey, right?

Harpreet: [00:59:41] Yeah. I've been trying to get Zack on my show. He's resisted all attempts to of my wooing him. He's just giving the cold shoulder.

Daliana: [00:59:49] When are you going to be on my show? I would love to have conversation with you.

Harpreet: [00:59:53] Yeah, absolutely, man. I mean, I'll be in California visiting family in December of this year. We'll see.

Daliana: [00:59:58] Yeah, maybe we can do a live recording. [01:00:00]

Harpreet: [01:00:00] Yeah, yeah, definitely. Yeah. We'll be in touch about that. Let's talk about your experience being a woman in tech and a woman in data. I'm wondering if you have any advice or words of encouragement for the women in our audience who are breaking into or currently in our data world?

Daliana: [01:00:15] Yeah, so I don't have a research data, but from my observation, I show that there are more women in data science than in engineering. So you won't be lonely even if you are. And that's okay. I think I my previous AWS VP says something I loved. She said, You know, you shouldn't wait to do something. You know, only when you see someone looks like you who have done it right. You can do anything if and you don't have to wait for like a permission to do something. Maybe you're the only woman in your company, and that's okay. When you go into a meeting, you don't have to always think about, Oh, I'm the only woman here that's just going to make you nervous to think about what's going to be your message, what's going to be your voice, what type of value are you going to provide for the meeting? And just because you're the only woman in the meeting, by the way, I happen to me a lot. It doesn't make me feel intimidated. It made me feel excited. Okay, I'm the only female here. That means I have to say something to represent my group, my perspective. Right? Maybe other people. I wouldn't say people would intentionally ignore women's voice, but maybe from their perspective they were not. Think about something that would impact the female users.

Daliana: [01:01:36] For example, if you're working on a product and definitely there, they're going to be there already a lot of communities for a female scientist or engineer, for example, the Grace Hopper Conference. And another thing I have, I think it hasn't been talked about, maybe a little bit controversial. I want to want to see what you think about it. I think one [01:02:00] thing is the important to go to those communities to get support, to see what other female scientists career journey. But also don't forget to go to those general conferences when it's not focusing on women, because that's what the reality is. You're going to work with everybody. So if you only feel comfortable working with women and then maybe in the real world, you're going to feel intimidated when you have to collaborate with a male collaborator. And also, I have great female mentors, but I also have great mentors, men that are men. So if you're doing, you only talk to women in size, you're limiting yourself for other resources. So I would say find your support system, definitely find a group that you feel comfortable. You can, you know, talk about all your struggle with specific for being a woman, but also don't forget to interact with other people.

Harpreet: [01:02:57] Absolutely love that advice. Thank you so much. Damien, I wonder what what can we do as as we like? Me and the other men in data science. Like what? What can we do to foster the inclusion of women in data science? And I.

Daliana: [01:03:11] Yeah, I think that's. Yeah. Thanks for asking that. We're just thinking about that. I think it's to when you have a model, when you're developing a product, just to think, oh, who's not in a room, right. How we include women or other minority groups in tech. I think just having that thought is is important also when it comes to mentorship know that bias. So if you if I look at my LinkedIn messages most of them from men you can say, you know there's a base rate there are more male scientists, engineers in the field. But also I feel guys are more likely to ask for help, you know, to demand things. And for women sometime where we don't want to income wins people so maybe at work if you see someone struggle [01:04:00] maybe she didn't ask you to become a mentor, but maybe you can offer to see, Hey, do you need my help? Do you have a mentor? And I think especially when someone join a team to to provide those type of help, whether you are male or female. Data scientists just know that there is this type of behavior differences. I wouldn't say it happens to everybody, but at least from my observation on my LinkedIn messages.

Harpreet: [01:04:29] Yeah, that's an excellent point. I was talking to Natalie Nixon. She's the author of The Creativity Leap. And we're talking about she was telling me a story about how like back in back in the days in school, like in physics class, like any time there's a question to be answered, like boys hands would go up and they'd all have the wrong answer. But they wouldn't they wouldn't be afraid to just shout out that answer, whereas women would be more kind of reserved and wouldn't want to shout out the answer. And she's talking about the importance of just just putting yourself out there. Just do it like, you know, there's no especially when the downside is so limited and the upside is unlimited. Thank you very much for that. Let's go to the last question before we go on to a real quick random round. It is 100 years in the future. What do you want to be remembered for?

Daliana: [01:05:17] Yeah. I think after 100 years are probably just to want people to forget about me. But I think I get the gist of your question. What would be some work that I really want to make an impact? I think I still really enjoy data science. I think data science is really cool. It can help us with everyday decision making. I really want to create something that make it easy for everybody to understand the data science concept. Maybe I would. Don't be surprised if one day I write a cartoon book or tell stories with data science. I just think that's something really cool if people read my data science adventure stories. But another thing [01:06:00] is I want people to feel seen in their career journey, and every time I share some struggles, people will say, Oh, you're exactly talking about what happened to me. I think I want people to feel they're not lonely in this career journey. I think now with LinkedIn is a community and also with a podcast. I don't know what I'm going to do in the future. I think maybe I can create something that allow people to support each other on their career journey and be authentic. And maybe people can just remember that, you know, some time ago there was this data science called Dolly something, something, whatever. And she's a data scientist. She helped people use data scientists to make better decisions. And she really enjoyed her life and have fun and, you know, make other people have fun where a well enjoying their career and the life.

Harpreet: [01:06:55] I actually love that man. I'm looking forward to a Dilbert esque data science type of comic strip coming out from you. That would be really cool, man. Yeah. Let's go ahead and jump into the random round. We'll ask a couple of questions and then we'll go to a random question generator. In your opinion, what do most people think within the first few seconds of meeting you for the first time?

Daliana: [01:07:19] Yeah, people probably think I'm extrovert, but actually I'm kind of 50% introvert any time to be alone. So sometimes I do have to explain that to them.

Harpreet: [01:07:30] Thank you very much. So an introvert, it's it's a challenge. It's a.

Daliana: [01:07:35] Struggle. Like I need my time to write and think about stuff and read books.

Harpreet: [01:07:41] So, yeah, you do like journaling in the morning or anything like that?

Daliana: [01:07:44] Yeah, in the evening. And usually Friday. Saturday I write. Write down some ideas that I write down during the week for LinkedIn post the week after.

Harpreet: [01:07:55] I love that. Speaking of reading, what are you currently reading?

Daliana: [01:07:58] Currently, I joined the new [01:08:00] team actually recently and reviewing some forecasting books and old regression concepts and I'm rereading how not to be wrong. That's one of my favorite. I wouldn't say like data science. Data science. See, books I really love, books like that make you have a deeper understanding of the concept through stories I highly recommend. I think you interviewed the author, right, Jordan?

Harpreet: [01:08:25] Yeah, yeah. Jordan Ellenberg had him on the show. How not to be wrong how mathematical thinking that an extra copy of his. They sent me prerelease copies of shape. If you want one, I'll send one to you.

Daliana: [01:08:36] Wow.

Harpreet: [01:08:37] Yeah, I'll definitely get one out to you. Speaking of books, man, two too many books. Too little time. I start from this problem. Can you share just a couple of tips on how not to feel bad not finishing a book?

Daliana: [01:08:48] Yeah, I think it depends on what our goal is of reading the book. And a lot of times the authors write something because they think it's important in the system, but it might not be useful for you. So when you pick a book, I think just go into some area where you're most interested in, learn something immediately. You already get some ROI from the book. And then if you feel you need to learn a topic in a better system systematic way, then you can go through the book. Otherwise, if you just think about it, you read a quote from LinkedIn you wouldn't feel, Oh, I didn't know all the things this person said. You still learn something from the quote, so just use the same mindset when you read a book and also know, okay, maybe for the next year my focus is going to be about learning how to write. Then I'm going to write three books about writing. Those are the books maybe I do want to finish or a few chapters, but I'm also curious about biology. But that's not my priority. So I might still skim through some biology book, but just for fun. So have that priority in my mind. And when you're not finishing [01:10:00] the book that on your secondary category and you wouldn't make it too harsh for yourself.

Harpreet: [01:10:07] Go ahead and go to a random question generator. Thank you for sharing those tips as well. You can find more of those tips. Diane has video all about how not to feel bad about reading books. All right. We've got the random question generator up and running. First question is pirates are ninjas.

Daliana: [01:10:24] Pirates are ninjas. Pirates? I'm a sea person. There you go. Beyond water.

Harpreet: [01:10:29] Yeah. Mountains or ocean?

Daliana: [01:10:32] Ooh. Can I have both? I'm probably ocean. Oh, gosh. Yeah.

Harpreet: [01:10:38] If you were a vegetable, what vegetable would you be?

Daliana: [01:10:43] I would be I would be a watermelon.

Harpreet: [01:10:47] Nice last one here. If you could live in a book, TV show or movie, what would it be?

Daliana: [01:10:54] Oh, that's such a great book. A great question. Do I have a favorite movie, TV show? I don't know what I have been watching recently. I think I maybe want to. Living. Like some history fiction book. I don't remember the name of it, but I just feel it has some truth in it. But it's also this creativity element in it. And sometimes I think about if I go back in time, what would I create? Something like that. What about.

Harpreet: [01:11:29] You? Well, I've been really into this show called Foundations. I think.

Daliana: [01:11:33] Oh, I've been watching that to live there. But it's like, you know, the world is ruled by clones.

Harpreet: [01:11:39] And yeah, it just it seems interesting just to be in a place where we know that life exists and we're all interconnected in some type of galaxy. Like, yeah, I mean, genetic empire thing. Like, it's kind of weird, but that's still at the same time, it's actually kind of cool. But yeah, it'd just be, it'd be cool. Like, I'm big into, like, space movies and.

Daliana: [01:11:59] Stuff. [01:12:00] Oh, cool. Yeah, yeah.

Harpreet: [01:12:02] So, Eliana, how can people connect with you? Where can they find you? Online?

Daliana: [01:12:06] Yeah, you can follow meeting Darlene Liu and you can also join my newsletter Daily and Hulu.com and of course listen to my podcast, the data scientists show. And maybe that's too much, Diana, for you. Just pick a few to follow. Consume with caution you might get addicted.

Harpreet: [01:12:26] Yes, you definitely it. I'll go ahead and be sure to link to all that in the show notes, guys.

Daliana: [01:12:30] Thank you.

Harpreet: [01:12:31] I'm sure most people already know you and appreciate all your content as much as I do. I thank you so much for taking time and your schedule to be on the show today. I appreciate having you here.

Daliana: [01:12:39] Yeah, thank you.

Harpreet: [01:12:41] My friends, remember, you've got one life on this planet when I try to do some big cheers, everyone.