SDS 493: Bringing Data to the People

This is episode number 493 with Anjali Shrivastava, data scientist, artist, and producer of the Vastava YouTube channel. 

Jon: 00:00:12

Welcome to the SuperDataScience Podcast. My name is Jon Krohn, a Chief Data Scientist and bestselling author on deep learning. Each week we bring you inspiring people and ideas to help you build a successful career in data science. Thanks for being here today, and now let’s make the complex simple. 

Jon: 00:00:42

Welcome back to the SuperDataScience Podcast. You are in for a treat today with a fun, intelligent, and vibrant Anjali Shrivastava. Anjali is an expert in data science visualization. She has used this skill set to engineer visualizations of data in production systems in a number of roles. We’ll talk about that technical content in this episode, including her favorite tools and applications. But we’ll also discuss in detail Anjali’s mission to bring a face to data, which she accomplishes through journalism, as well as through her brilliant YouTube channel focused on applying data analytics to pop culture themes. 

Jon: 00:01:19

Anjali holds dual degrees from the prestigious University of California, Berkeley, in data science, as well as in industrial engineering and operations research. A recent graduate, she’ll fill us in on what a data science degree curriculum is like at a top university like Berkeley, as well as how anyone can access their world class data science lectures online. Today’s episode is well suited to listeners at any stage in their data science journey, whether you’re considering a move into the field, have recently become a data professional, or are a seasoned pro. All right, you’re ready for another awesome episode? Let’s go. 

Jon: 00:02:00

Anjali, welcome to the show. It is so great having you here. How you doing? Where in the world are you? 

Anjali: 00:02:07

I’m doing great. Thank you for asking. I’m currently in Sacramento, California. 

Jon: 00:02:12

I guess you’re from there, but you studied at Berkeley, which is how far away is Berkeley from Sacramento? 

Anjali: 00:02:20

It’s about two hours on a good day, traffic permitting, not too bad. 

Jon: 00:02:28

And so, you’re you’ve just finished a bachelor’s degree at Berkeley, and the pandemic has ruined your educational experience, I guess, your senior year. 

Anjali: 00:02:40

A little bit. Yeah, you could say that. Well, I finished two degrees, actually. 

Jon: 00:02:44

Wow. 

Anjali: 00:02:46

Yeah, a Bachelor of Arts and a Bachelor of Science. Arbitrary decision that they make, but yes, so I have a Bachelor of Arts in Data Science because data science is an art. And I have a Bachelor of Science in Industrial Engineering and Operations Research. 

Jon: 00:02:59

Wow. That is a really cool degree. 

Anjali: 00:03:02

Thank you. 

Jon: 00:03:03

Or set of degrees. Ah, I did it again. 

Anjali: 00:03:05

Thank you. Thank you. I mean, yes, there is quite a bit of overlap between the two. So, I think it’s fair to say I did get a degree’s worth of education, and I just decided to take on multiple classes. But yes, the pandemic definitely was challenging in a lot of ways, particularly in online learning. I think one thing that I say a lot is I was a senior, so I was already going to be a bad student because of [crosstalk 00:03:31], but online learning just made it doubly. So, it was a very, very hard to stay focused, even with the amazing educators at Berkeley. So yeah, a bit of a bummer, but… 

Jon: 00:03:44

Well, I wasn’t even thinking about the education part of being in university. 

Anjali: 00:03:47

That’s true. 

Jon: 00:03:48

I was just thinking about how university is this once in a lifetime experience to meet people and have fun and you can’t really do that over a Zoom? 

Anjali: 00:04:00

Oh, yeah, definitely. Yeah, no, we missed a lot out on that. Particularly, the fact that my 21st birthday was about a month after university went on to lockdown so we missed out on that, as well as every single one of my roommates and fellow friends in college, we were all turning 21 in that year. So, it was just a string of disappointments, but I mean, yeah, in the grand scheme of things, all of that doesn’t really matter. I’m still very happy that I got to finish out my degree, and I still think that I got a worthwhile experience out of it. I think learning through these remote challenges… I don’t know something everyone had to [crosstalk 00:04:37]. I think Berkeley made the most of it. Especially with our commencement, I think they did a lot of nice things. 

Jon: 00:04:43

That’s great. Your commencement was virtual as well? 

Anjali: 00:04:46

Oh, yeah. I think I actually had four different commencements just because of again, my two different degrees. 

Jon: 00:04:50

Your arts, your science degree. 

Anjali: 00:04:53

Right. Yeah. I think it was really sweet. They had our own slides. They read out our names, which they typically don’t do at commencement. My family in India was able to join. So, it was actually nice that would be possible otherwise, so yeah. 

Jon: 00:05:06

Now, for our international listeners, you might be wondering the significance of the 21st birthday in the United States. In the United States, the drinking age is 21, which is crazy. I don’t know. It’s draconian. And in other countries, in the UK, I think most European countries it’s 18 or younger. In Canada, in most provinces, it’s 18. A couple are 19, and it’s crazy you should be 21 before you can legally drink. Anyway, so yeah, too bad that you weren’t able to do that with friends there, but I’m sure the people you were close with you still, you can meet up with in real life now. 

Anjali: 00:05:48

Yeah. And I had, my dad and I, we went to Safeway. I bought my own can of cider for the first time. 

Jon: 00:05:54

[crosstalk 00:05:54] the grocery store. 

Anjali: 00:05:54

It was exciting. Yes. 

Jon: 00:05:55

That’s funny. 

Anjali: 00:05:57

Just a few blocks away from my house. It was a nice little walk. Perfect celebration. 

Jon: 00:06:04

Cool. Well, so I should explain to the audience how we know each other. 

Anjali: 00:06:09

Yes. 

Jon: 00:06:12

So, we’re filming early July, and a couple of weeks ago someone named Harpreet Sahota, and someone else named Kate Strachnyi, both of whom have been guests on the SuperDataScience show this year. So, Harpreet was episode 457, and Kate was Episode 441. They hosted the first I assume annual Data Community Content Creator Awards. And Harpreet in the run up, he made a post saying, “Hey, this award show is coming up, and the winners are determined by voting.” So it’s a fan favorite voting thing, kind of like the People’s Choice Awards. And he specifically mentioned a number of amazing data content creators that you might want to consider voting for that he would certainly consider voting for himself. 

Jon: 00:07:10

And so, I went through all of the names in this list, and I came across you. So I go through to your LinkedIn profile. I go to Anjali’s LinkedIn profile, and it looks so interesting. So we’re going to get into all of this, but Anjali has this journalism background that she’s pursuing, at the time were pursuing now have completed dual degrees in data science and industrial operations- 

Anjali: 00:07:40

Industrial engineering, and operations. 

Jon: 00:07:41

Industrial engineering, and operations. 

Anjali: 00:07:41

It’s kind of a mouthful. 

Jon: 00:07:46

And had experienced early career experience in data science roles, data visualization, and I thought, “Wow, Anjali would be a really cool guest to have come on, and explain this early career journey.” And we do have tons of interesting content. I have so many great questions about this early career journey that I think the audience is going to love. But remember that Anjali was originally referred to me indirectly by Harpreet mentioning that she’s an amazing content creator. So, on her profile, I can’t really figure that out. And so, actually, I messaged Harpreet, and he pointed out to me your YouTube channel. So, Anjali has a YouTube channel. It’s the back part of her last name, Vastava, which actually explain that to us, I think it’s really interesting. So, your last name is Shrivastava. Your YouTube channel is Vastava. Explain what those words mean [crosstalk 00:08:41]. 

Anjali: 00:08:41

Sure. Yeah. I mean, so the original impetus for is Vastava was just my GitHub username. So, initially, I didn’t put a lot of thought into it. It was just it was there for the taking. But I came to realize after talking with my dad oh, Vastava is actually the Sanskrit translation of either truth or fact, which really aligns with what I’m doing. It’s tangentially related to data science, this idea of objective truth. And then I was thinking about it more like Vastava was always something that I had in my mind because when I was younger I would often cut off the Shri in my name because Shri is an honorific. It’s loosely translated as Mr. So, my last name could be thought of as Mr. Vastava. And I used to be like- 

Jon: 00:09:25

Mr. Truth. 

Anjali: 00:09:26

Yeah, Mr. Truth, a great superhero name, but doesn’t really fit me. 

Jon: 00:09:30

You can’t lie to him. 

Anjali: 00:09:33

Exactly. So, I cut off the Shri because I’m not a Mr., obviously. So, I just thought it was funny that those three methods or explanations for the name coalesced into this one word that I think a lot of people when reading it just think it’s random syllables. But no, it actually has three different meanings to it if you think about it. 

Jon: 00:09:55

Yeah. Eliminating unnecessary distractions is one of the central principles of my lifestyle. As such, I only subscribe to a handful of email newsletters, those that provide a massive signal to noise ratio. One of the very few that meet my strict criteria is the Data Science Insider. If you weren’t aware of it already, the Data Science Insider is a 100% free newsletter that the SuperDataScience team creates and sends out every Friday. We pour over all the news and identify the most important breakthroughs in the fields of data science, machine learning, and artificial intelligence. The top five are simply five news items. The top five items are handpicked, the items that we’re confident will be most relevant to your personal and professional growth. Each of the five articles is summarized into a standardized, easy to read format, and then packed gently into a single email. 

Jon: 00:10:58

This means that you don’t have to go and read the whole article, you can read our summary, and be up to speed on the latest and greatest data innovations in no time at all. That said, if any items do particularly tickle your fancy, then you can click through and read the full article. This is what I do. I skim the Data Science Insider newsletter every week. Those items that are relevant to me, I read the summary in full. And if that signals to me that I should be digging into the full original piece, for example, to pour over figures, equations, code, or experimental methodology, I click through and dig deep. So if you’d like to get the best signal to noise ratio out there in data science, machine learning, and AI news, subscribe to the Data Science Insider, which is completely free no strings attached at www.superdatascience.com/dsi, that’s www.superdatascience.com/dsi. And now let’s return to our amazing episode. 

Jon: 00:11:57

So I checked out your YouTube channel and it’s amazing. I absolutely love the content. I had to get you as a guest on the show, reached out to you on LinkedIn, and was delighted that you are keen to be on. I think this is going to be such an interesting episode. So some of the more popular videos on the channel are TikTok mansions during the pandemic. So, it seems like you’ve found this niche of social media drama, as you described it. 

Anjali: 00:12:28

Yeah, doing data analysis of social media drama, important caveat, I’m not just talking about it. I’m applying my background in data science. 

Jon: 00:12:35

Yeah, so lots of grabbing data where you can get it, putting in charts, you make tons of custom graphics, and you overlay them. And these are long, detailed videos, which I think is great. Typical run times 10, 15, 20, 25 minutes, and you can really dig into the data around pop culture events. So, this TikTok mansions one I know is big. One that you clearly put a lot of effort into and is a lot of fun to go through is a Star Wars Legends versus Canon with data science. What does that mean? 

Anjali: 00:13:15

Yeah, so the Star Wars universe has two separate timelines. So, before Disney acquired Star Wars, there was this massive sprawling universe of books, comics, radio specials, TV shows, movies, that was called the Legends timeline. Or it wasn’t called anything, really, it was just called the timeline. Then when Disney acquired it, they decided to scrap all of that and they named it as the Legends timeline. And then they establish their own Cannon timeline, and it made a lot of sense why they did that. When they wanted to make their new movies, they could have a clean slate and not have to be tied down by all of this backstory. But what I ended up doing, which I did this crazy intensive project where I scraped every single event on the Star Wars Wiki pages of every event that happened in the Legends timeline and every event that happened in the Cannon timelines, and I would compare the two to see how different are they? There was a lot of interesting insights in that, and I put a lot of time into that. 

Jon: 00:14:13

Yeah, it’s clear. I’ve also seen some visualizations of it. So your GitHub account, vastatva.github.io, audience members can go there, and a lot of the visualizations that you use in your YouTube videos are there as standalone visuals. And the timeline that you put together, and the graphic you put together for the Star Wars video is epic. 

Anjali: 00:14:40

Thank you. Yeah, it’s still an interactive website. You can scroll lots of fun things pop up. There’s little light saber switches that I coded using CSS. Yeah, it was all coated in D3, the interactive data visualization library. I’ve been having a lot of fun with it lately. 

Jon: 00:14:56

Nice. Yeah, D3.js is a JavaScript library. So, yeah, so you obviously… Man, I guess we should dig into this a little bit now. You’ve had a couple of roles as a data visualization engineer. And so, obviously, in that role, oh, man, I guess we should talk a little bit about it now because it’s just going to be a tease to the audience if we don’t. So when you’re in a role, a data visualization engineer, evidently, and this makes a lot of sense to me, now that I think about it, you need to develop some expertise in things HTML and CSS and D3. So yeah, so maybe the mm-hmm that you just gave me should be enough for now. We’ll come back to those roles and talk about the specific tools and techniques that you specialize in. But we’ll wrap up with YouTube. I love this angle of data analysis of pop culture. The videos that you make are so fun, and relatable, but also there’s some technical elements. Huge amounts of effort go into making these. So anyway, yeah, really glad that you’re here on the show. And you even do have a couple of data science videos. So, you have a couple of videos on your YouTube channel that are not just analytics, but even some data science. You have a music genre classifier video. What is that all about? 

Anjali: 00:16:16

Yeah, yeah. So there’s this very famous internet persona, [inaudible 00:16:22] musician named Corpse Husband and everyone was freaking out about his music, but no one could really fit what genre it fit into, basically. So, that was the perfect… I was like, “Oh, this is perfect.” It’s an internet story, which is most of my videos. And also I can make an AI classifier of music genres, which is a classic project that I think a lot of us have done at some point. So that is really, I think, honestly, it’s the only video I have on my channel that uses machine learning in any capacity, which is surprising, because I think a lot of us when we think of data science projects, we immediately go to AI, ML. But really what I’ve tried to do with my channel is stay grounded in data analysis, data scraping, concepts that are more easily explainable to people who might not have background in science because we have to be honest, the vast majority, people don’t have a background in data science or basic statistics at all. So, I have to make sure what I’m doing is not only technically robust, but it also can be comprehended by everyone, which is why I think I would lean towards doing visualizations because I think visualization is the way that you can communicate data science findings to anyone because everyone can see. 

Jon: 00:17:33

Totally. It makes perfect sense to me. I love that angle. Something that we were talking about before the show is that as your data science career develops there will probably be opportunities where you see machine learning things, data science things that you’re doing at jobs where you’re like, “Wow, this is really interesting. I got to make a video about this.” Actually, that brings me to a question that I’m really curious about. So how do you get your inspiration for particular topics? How do you decide where you’re going to do a video on? There’s a huge variety of topics covered here. 

Anjali: 00:18:09

Yeah, I mean, I have a long list of ideas, and I pull from every experience. Like whenever I’m watching a movie I’m like, “Oh, that’s interesting.” So, like the TikTok mansions video that you mentioned. That was only because I was addicted to Twitter, and I’m sure people know Taylor Lorenz. She’s a New York Times reporter who her whole beat is covering influencers and TikTok and those things. I followed her. Literally, it seemed like every day there was a new mansion popping up. And I was so confused because I was like, “Why are people doing this?” Literally, in the heat of the pandemic. 

Jon: 00:18:45

What is it? You need to explain what to us. 

Anjali: 00:18:46

What is it? That’s true. So, TikTok mansion, excuse me, is when a bunch of teenagers move into a mansion together to create content together. There’s a bunch of different names for it. I think it’s also called a content house, which might make more sense to the mind. Basically, a bunch of famous people from all across the country will move together. And so, they can more easily collaborate with one another, and create TikToks together. The problem with it is, this was obviously happening during the pandemic. And also, these are a bunch of teenagers who are not necessarily the most responsible and have come into a lot of wealth recently. So, there were a lot of videos [crosstalk 00:19:20]- 

Jon: 00:19:19

Right, what could go wrong? 

Anjali: 00:19:22

Exactly. So, there’s massive parties- 

Jon: 00:19:23

You got a bunch of teenagers, unexpectedly very wealthy and famous, living together in the middle of a pandemic. 

Anjali: 00:19:30

Yes, without parents, right? 

Jon: 00:19:32

Oh, my goodness. 

Anjali: 00:19:33

Often and most of these were in LA, a city where none of them have grown up. So there are all these videos of these massive parties and it got really big. I think one of them the governor, or the mayor, excuse me, the mayor of LA had to shut down the water and the electricity at the house because they were just too much of a public nuisance. So, there was a lot of interesting stories to be told there, and I just wasn’t seeing that angle a lot on the internet. So, I decided to put my data science hat on. Not just my data science hat on, but I put my geographer hat on, because that actually… That was my concentration, GIS and technologies within the data science major. 

Jon: 00:20:08

Oh, no kidding. 

Anjali: 00:20:10

Yes, it was. I really love maths and random geography skill. 

Jon: 00:20:14

What’s GIS? Geo Information Systems? 

Anjali: 00:20:18

Yeah, Geo. 

Jon: 00:20:18

Geographical Information- 

Anjali: 00:20:18

Yeah, Geo Information Systems, or Geo Information Sciences are the two acceptable acronyms. But yeah, I made this interactive map. It’s also on my website, if you want to check it out. You can scroll. It’s super interactive. You can click the dots. I’m actually quite proud of that one, also. 

Jon: 00:20:36

Have you ever had one of your YouTube projects overlap with an academic project? 

Anjali: 00:20:47

I can’t say that I have because I started my YouTube channel in summer when I wasn’t taking classes, so that wasn’t really possible. And then in my senior year, I don’t know, I think a lot of my projects, it didn’t work out that way, but [crosstalk 00:21:04]- 

Jon: 00:21:03

So, you didn’t write a paper for your GIS degree about TikTok mansions? 

Anjali: 00:21:08

I did not, unfortunately. I think my professor would have laughed me out of the room if I did, but… Maybe. 

Jon: 00:21:18

All right. So, that’s really cool. I’m sure we’ll end up talking about the YouTube channel again. Lots of interesting aspects to you, Anjali. Another one that we talked about a little bit before the show started was that you are into yoga, and it was interesting. So you mentioned to me that through the pandemic, you fallen off a yoga wagon, that you’ve lost your routine around that. And that’s interesting to me because in episode 473 with Anima Anandkumar, who is prominent machine learning researcher. She’s at Nvidia. She is a professor at Caltech. She had the opposite experience. So, she grew up in… Oh, my goodness, I can’t believe I’m blanking on the name of the region in India that modern yoga comes from. 

Jon: 00:22:11

Anyway, it’s embarrassing that I can’t remember it, but she’s from that region. There’s all kinds of listeners shouting it at their podcasting app. Anyway, what can I do? It’s life. So, despite growing up in that region of India, she had never really taken up very much yoga. But when the pandemic happened she told us in the podcast episode that she got into it. And then it became a really great way of centering, and dealing with this weird experience of the pandemic. I think, especially, at the beginning of the pandemic was so bizarre. Now we’re just used to it. I’m indoors all day. That is life. But yeah, but you went the other way. 

Anjali: 00:22:57

Yeah, no, the reason I fell off is, well, first of all, I was really into yoga at Berkeley, and there were a lot of classes that I would attend that were free through the Cal ID. So, there were a lot of very talented instructors at the gyms that I formed relationships with. So obviously COVID did not permit that to continue. But the second probably more important reason I stopped doing yoga is we actually got a puppy during quarantine, and he’d just been consuming all of my time. So, having a puppy honestly requires a lot of cardio exercise every day because he just is this ball of energy even now 12 months later. He’s one-year-old now. So, yeah, and there’s a couple of times when mum and I have tried to sit down and do yoga and you just can’t do it without the dog putting his face in and wanting to be a part of it. 

Jon: 00:23:51

You could only do two poses, down dog and puppy pose. 

Anjali: 00:23:53

Yeah, literally. 

Jon: 00:23:57

So, all right, let’s talk about the puppy for one second. What kind of puppy is it? 

Anjali: 00:24:00

Oh, sure. He is a black Lab. His name is CJ. He is quite big. 

Jon: 00:24:04

Lots of energy. 

Anjali: 00:24:05

Yes, he is almost as heavy as I am, actually. He just passed 97 pounds. 

Jon: 00:24:10

Wow. 

Anjali: 00:24:11

So, he’s a big one. Yeah, definitely not able to carry him anymore. 

Jon: 00:24:16

That would be incredible. 

Anjali: 00:24:19

Yeah. 

Jon: 00:24:20

All right. So super interesting. I’d love to learn these bits about your personal life. But I do also want to get deep into the data science. So let’s do that by talking about journalism. 

Anjali: 00:24:35

Okay. 

Jon: 00:24:38

So, you actually, so this is going to make sense to the listener momentarily. But Anjali identifies as a writer and you started your career in journalism. So, I guess one of your first jobs was at the Daily Californian, which first of all, you’re going to have to tell us a bit about that. I guess it’s a newspaper. And so, at the Daily California, you were assistant news editor, and then later senior staff reporter, so tell us about that. 

Anjali: 00:25:09

Yeah. So, I mean, I would hesitate to call it a job even though… So, the Daily Cal is a student news room, first of all, I should clarify [crosstalk 00:25:18]. I will say it was very demanding, and definitely on par with the job. But yeah, I just don’t think that terminology is necessarily accurate. But the fact that it’s a daily newsroom, which means we are pumping up stories every day. So particularly when I was an editor, it was a six day a week job. I was paid for it. So, yes, it was a very intense time of my college career, and I learned a lot there. And that newspaper has churned out a lot of very successful journalists for good reason. There’s really no better training than being on the clock 24/7, especially in a place Berkeley, which is both the city and the university. There’s so many big things happening every day that it’s really impossible to not find those big news stories. You could just trip and you’ll find some big protest, or some lawsuit involving a professor or something. It’s just a campus full of very notable people, so there’s always a lot of notable stories, I think. 

Anjali: 00:26:19

So that was really a formative part of my college career in the sense that I learned how to navigate these very big stakes situations, and also be comfortable talking to people who have wildly different backgrounds and stories and perspectives. But yeah, I know, the reason I say I’m a writer first in my bio is because even though… The way that journalism and data science intersect, for me at least is when I’m doing data analysis, I’m always trying to find the story in the data, trying to put together a narrative that makes sense with the analysis. And I think that leads really nicely into my YouTube channel. I’ve always been really interested in not only finding the questions that can be answered with the data, but why does that question matter, and what does that mean to people? 

Anjali: 00:27:05

So, I think, in data science, a lot of the times the story… Not these stories, but our analysis, and our projects can come off as very clinical. And I think there’s been a push in academia recently to make ourselves recognize those more human elements or inject some sort of emotionality into it, whether that’s through interviewing people, or just doing background research on the people that you’re analyzing. And that’s probably more relevant in demographics than it is on doing a machine learning project for whatever tech company, but I think that skill of being able to put together a story whether it’s through hard data, or whether it’s the facts of an interview and parsing through facts that you find from journalism. That’s a very important skill for almost every one, and particularly for data scientists. 

Jon: 00:27:58

Even with a lot of machine learning algorithms, you’re building them for people, they impact people. And we probably should be as data scientists spending more time getting to know the end users who in a lot of cases could be very different from us. 

Anjali: 00:28:18

Right. I think my goal with my YouTube channel was to put a face to the data. Even though the ideas and projects I’m analyzing are very low stakes, the grand scheme of things, and that’s by design. I’m intentionally picking low stakes things because I know I don’t have peer-review processes. I don’t have anything. This is just me in my bedroom doing random analysis projects. No one’s going to really care if I screw- 

Jon: 00:28:42

A YouTube channel doesn’t have peer review. 

Anjali: 00:28:45

I do not, unfortunately, just a 22-year-old in my parent’s bedroom. But yeah, so I mean, sometimes I think I do have a video on COVID. I do have a video… I had a very fun video making a drinking game for the presidential debates right before the election using TF, IDF, and those sorts of things, but- 

Jon: 00:29:09

No kidding. 

Anjali: 00:29:09

Yeah, that was- 

Jon: 00:29:11

So, TF IDF, term frequency, inverse document frequency is a way of identifying the most likely relevant phrases- 

Anjali: 00:29:19

Unique phrases. Right. Yeah, common phrases that aren’t just the AI, etc., unique and common. I’m impressed that you knew what that stood for. I was like, I totally remember. 

Jon: 00:29:31

I work on natural language models, I got to know. Yeah. 

Anjali: 00:29:37

Good to know. Okay. So, yeah, that was… So, I guess those videos can be considered high stakes. But yeah, even for low stakes projects like the ones I’m doing I think it’s valuable to have someone who knows the inside and can explain in English language to everyone or I shouldn’t say English, in language like I do so most people can understand. 

Jon: 00:30:02

Yeah, in natural language as opposed to zeros and ones. 

Anjali: 00:30:05

Mm-hmm (affirmative). Having a human interpreter for the data. 

Jon: 00:30:09

Yeah. Very cool. I really that idea of putting a face to data. And you’re definitely doing that with your YouTube channel. I think it also especially appeals to probably generally a younger audience that might- 

Anjali: 00:30:22

You’ll be surprised, actually. 

Jon: 00:30:22

Oh, yeah? Okay. Do you have a sense of your demographics, your audience demographics? 

Anjali: 00:30:27

Yeah, yeah. So, my biggest demographic is 25 to 35, and then my second biggest is 18 to 25. 

Jon: 00:30:36

To you those are very old people. Could you imagine people as old as 35? 

Anjali: 00:30:41

When you hear young people on YouTube, my mind goes to 11 to 14 year olds, and those people are not… Those kids are not going to be watching my content. But yeah, so I guess young people in the sense of people my age, yes, 18 to 35. 

Jon: 00:30:57

Very good. Yeah, on the younger end of an entire lifespan, but maybe not for you. 

Anjali: 00:31:05

Yeah, young in the sense, in terms of career, but not… Because when I hear young, I really think YouTube Kids. I’m not producing content for the kids. 

Jon: 00:31:15

I understand. So, you were working at the Daily Californian while you were at UC Berkeley, I guess? 

Anjali: 00:31:20

Yes. 

Jon: 00:31:21

Okay. So we’ve already talked about the UC Berkeley degree a bit, but very interesting. Bachelor of Arts in Data Science, Bachelor of Science in Industrial Engineering and Operations Research. So, I think we should talk about both of those and let people understand. So first of all, obviously, data science podcasts, people are going to be interested in knowing what a BA in data science is all about. When you came to Berkeley, did you know that this was what you wanted to study? 

Anjali: 00:31:52

Absolutely, not. I came into Berkeley undeclared. I was throwing around a lot of different ideas. I was thinking maybe economics or applied math was one thing that was really circling around. In my freshman year, I happened into a data researching role at the D Lab at Berkeley. So, the D Lab it’s short for the data lab. It might sound like a data science lab. It wasn’t necessarily that. The D lab is primarily designed to help social science researchers do data analysis to complement their reports. So it’s like you had enough professionals collaborating with more humanities, social science [crosstalk 00:32:30]. 

Jon: 00:32:30

The social science people need lots of help. [crosstalk 00:32:33] adults in the room to help= 

Anjali: 00:32:34

Oh, I probably didn’t phrase it [crosstalk 00:32:38]. But yeah, I guess it’s funny from the beginning of my career I was always thinking about how data and writing can mesh. But yeah, so the role was nothing fancy. I wasn’t necessarily learning Python, or R, or anything. It was mostly data collection, and having to think about data and what format data should be in order to be analyzed properly. And lots of questions you’d ask, but not that interesting. 

Jon: 00:33:03

So, kind of how surveys are designed, and how we should store the data, these things should be integers, and these other things should be [crosstalk 00:33:15]? 

Anjali: 00:33:15

Yeah, so because the nature of the D lab, the data that we were getting, it wasn’t what you’re thinking of. It’s not a CSV. We would be getting historical documents, and that would be [inaudible 00:33:24]. People would read them and tag them in certain ways or they’d have annotations. It’s like, “Okay, how do you convert these annotations into data?” So, normally, there would be a data scientist that we’d be working under and we think, “How can this best be translated for a data analysis or data visualization?” And I got those- 

Jon: 00:33:43

So like, a social scientist comes in and circles a bunch of words on a bunch of pages, and then brings it to you, and it’s like, “I know there’s something here, but I don’t know how to record this.” How do I convert this into something that I can put into a spreadsheet or something? 

Anjali: 00:33:58

Yeah, yeah. And that’s probably not the fairest characterization. The social scientist is usually very involved. 

Jon: 00:34:05

I know. I’m just picking on the, 

Anjali: 00:34:08

I know. Yeah, I’m too cautious. [crosstalk 00:34:11]. 

Jon: 00:34:13

I love all your [crosstalk 00:34:13]. I’m just having a bit of fun. 

Anjali: 00:34:13

Yeah, ditto. But yes, the data scientists and social scientists are usually leading a project and leading the researchers to think about these questions, and think about what questions can be answered with the information that we’ve been given. I found those challenges, I guess, very interesting. I thought it was like a puzzle, and I thought I was really good at it, honestly. So I right then and there I was like, “Oh, maybe I should look into data science.” The problem was when I was a freshman the data science department and the data science degree did not exist at Berkeley. So the closest analogs were either computer science statistics, or what I ended up choosing was industrial engineering and operations research. I was really drawn to that because… So the idea industrial engineering and operations research is essentially optimization. How can you optimize processes? How can you optimize shortest path? Basically, it can be applied in a lot of different routes and industries. 

Jon: 00:35:15

I can imagine. It sounds hugely valuable to almost any organization. 

Anjali: 00:35:20

Right. 

Jon: 00:35:21

Not just the industrial processes, but commercial ones, and yeah. 

Anjali: 00:35:25

Yeah. And as someone who was really confused, and didn’t necessarily know, even I knew I wanted to work with data, but I don’t necessarily know when that will take me, IOR, as it’s called seemed it could lead me to a lot of different places. So I ended up going with that degree. And then, after my sophomore year, which was two years at Berkeley, the data science degree was announced. So then I was actually able to do both. 

Jon: 00:35:51

Nice. Very cool. So, what do you do in a data science bachelor’s degree, I guess, at Berkeley specifically, but just in general? So, what’s the foundational curriculum of a Bachelor’s in Data Science? 

Anjali: 00:36:10

Yeah. So the data science department at Berkeley is really like the marriage of computer science and statistics. [crosstalk 00:36:17]. So, yeah, we have a couple of classes within the data science department that are required, but the vast majority of our classes are taken outside of it either in CS Stats or another department that’s related to your concentration, which for me, again, was GIS. So we have this fantastic course at Berkeley called Data 8 that’s actually available to anyone. Even if you’re not at Berkeley, you can take it online for free through edX. So, I do [crosstalk 00:36:42]- 

Jon: 00:36:41

The number eight? 

Anjali: 00:36:42

Yes, just look up Data 8 eight, you’ll find it. It’s an introductory course to Python and statistics and computer science. It’s everything you need to start. It’s kind of funny, even though that’s the intro class that you should take first, I didn’t end up taking it until my senior year just because I needed it to qualify for the degree. Even though I knew a lot of the stuff that was taught in the course, I had to take it just so I… So, yeah, I took the prereqs after I had taken everything else. That’s just because of the wonky way in which the degree was approved after I had already taken all these. Yeah, Data 8 and then foundational computer science, foundational statistics, and then with that you’re on your way to doing more upper division courses. 

Anjali: 00:37:27

So for me, just because I was part of that inaugural class. I think I was the second class to graduate, there weren’t many classes available when I was there. So we would end up taking a lot of graduate courses or again, classes that were tangentially related, but I was really lucky because I was an IOR, there was quite a significant overlap between the two tracks. So, I did, like in my IOR coursework, I learned a lot about stochastic processes, I learned a lot about optimization, nonlinear and linear. I learned a lot about there was a machine learning class I had to take for IOR, a lot of these algorithms that we use in machine learning. I look more how to derive them in my IOR. So, that informed a lot of the data science curriculum that actually wasn’t necessarily there at Berkeley. It was there, but it wasn’t part of the data science part yet. So, I think they’re still working out the kinks. But I mean, I still really thought most of the classes I took there were worthwhile, and there’s just so much available to you. I think the most overwhelming part of it is choosing which courses to take because there’s so many tracks you can go with it. 

Jon: 00:38:35

I am so jealous. I wish I could go back in time and be coming out of university with your education. That is such a cool curriculum. It’s like, these are things that I cobbled together year over year on the side while I’m working to have that background. It sounds amazing. 

Anjali: 00:38:54

I think one of the fantastic things about remote learning those now that the courses are online, if the university permits it, they’re really accessible to anyone. It’s not just the students anymore. So I remember at the beginning pandemic, I think we had to shut this down because it wasn’t exactly legal. But all of the students on most of these top universities were exchanging links with each other. There was this massive spreadsheet where we could just visit any class. You could hop into a Zoom at Princeton or Caltech, anywhere. And it was really interesting. I remember just watching a couple of lectures at other places, how different approaches at different schools. Yeah, it’s kind of [crosstalk 00:39:31]. 

Jon: 00:39:31

There are a lot of amazing… Yeah, university lectures, a lot of universities. It’s not every course but a lot of courses the lecturer says, “Yeah, I’m happy for this to go up on YouTube or inexpensive or free edX.” That’s what you mentioned, right? For the Data 8, was that edX 

Anjali: 00:39:46

edX, yeah. Yeah, it’s just a great resource. 

Jon: 00:39:49

E-D-X for you listeners out there. And so, a lot of what I know about deep learning and that is in my book, Deep Learning Illustrated, I got from YouTube lectures, from university lectures online. And we actually have coming up next month I am interviewing Professor Pieter Abbeel who is at UC Berkeley. And the way that I know him is he has on YouTube tons of deep reinforcement learning lectures. He specializes in robotics, and he’s got these this great entire course, many years of courses of deep reinforcement learning content. And I can’t believe that he’s going to be on the show, and I’m going to get to interview him. 

Jon: 00:40:33

Anyway, but so yeah, I’ve lost the point that I was trying to make. But it is cool that so many of these resources are available online for people to learn. So, people like me who I was already professional working done my PhD when the term data science was coined. So, obviously impossible for me to have a formal education in data science. But there are, as you say, so many cool free or very inexpensive resources out there for constantly learning. And it’s not like the learning ever stops in data science, right? It’s not like, “Oh, wow, because you have a degree in data science, you’ve got it all.” There are some things like knowing how Stochastic gradient descent works. Some mathematical concepts that won’t change, but there’s always new software libraries, new approaches. And so, you just got to get used to it. You got to listen to things the SuperDataScience Podcast to know what’s going on. 

Anjali: 00:41:34

Yeah. And I think what Berkeley does really well is early on they instill you the skill of reading documentation and figuring out new libraries. So, I think, once you have the foundational concepts of knowing how to code what functions are usually built in, then with that skill of being able to read documentation and knowing how to Google well, you’re really ready for almost anything, I think. 

Jon: 00:42:00

Nice. I’m going to go on and ask you a couple more questions about your data science background, but before we do that, I have a really important announcement for everyone, which is that I remember the name of that region in India where yoga is from and Anima Anandkumar is from. It’s, I’m probably going to mispronounce it, but Mysore, M-Y-S-O-R-E. 

Anjali: 00:42:21

Mysore? 

Jon: 00:42:21

Mysore. 

Anjali: 00:42:23

Mysore, yeah. 

Jon: 00:42:24

I told you I’d pronounce it wrong. Okay, so the D lab, Berkeley data researcher experience was critical for you as somebody who came into your degree, undeclared. And that allowed you to figure out that you wanted to be going to this data science thing. You had to have this interim degree. This Bachelor of Science in Industrial Engineering and Operations Research, which sounds like a really cool discipline. 

Anjali: 00:42:54

It is. 

Jon: 00:42:54

Hugely valuable in the world. And then going into your third year, this BA in data science comes out, and you get that degree as well. So while you’re pursuing this degree, you’ve had a number of really cool data science roles. So you were a business systems analyst intern at Intuit, which is a big software company. They make QuickView, is that right? 

Anjali: 00:43:18

They make on financial services. So, TurboTax is the one you’re probably most [crosstalk 00:43:24]- 

Jon: 00:43:24

Oh, TurboTax. 

Anjali: 00:43:24

But yeah, they also do Mint and QuickBooks is the other flagship service. 

Jon: 00:43:31

Oh, yeah. And then more recently, or maybe it isn’t something that you recently specialized in, but it is something that it seems you specialized in at least the role titles is you were a data visualization developer for the Berkeley Political Review. And then a data visualization engineer at HiGeorge. So I don’t think I’ve ever had anyone on the show who specifically even in their… So certainly, I’ve had data scientists on the show who have some expertise in visualizations, but I think you’re the first to have it in their job title. So, obviously, heavy expertise in that role. So what kinds of tools and techniques do you use in roles like that? 

Anjali: 00:44:21

Yeah, I mean, so I guess I should start off by saying at Intuit one of my big summer long projects was redesigning all of the most standard dashboards that people in the finance department use. So, for that, I ended up having to do interviews with a lot of people who use these dashboards and figure out what charts work, what metrics do they want, blah, blah, blah. And that started me on this journey of data visualization because it really got me thinking about how do you best display the information? Not just what information do we need to display, if that makes sense. 

Jon: 00:44:52

Totally. 

Anjali: 00:44:53

So after that work experience, I took a course in data visualization theory at Berkeley, and was also involved with Berkeley Political Review, BPR as you said. That, honestly, it was important for me to see how visualizations work alongside a story and how you can complement writing arguments with data. 

Jon: 00:45:15

What is BPR? 

Anjali: 00:45:16

Berkeley Political Review is a political magazine at Berkeley. I believe they came out once a month, so they would pair us with writers. The one that I work with a writer was writing about the Supreme Court and how the selection process is random. So, I had made a nice Tableau dashboard, I can send it your way if you’d like. 

Jon: 00:45:38

Cool. 

Anjali: 00:45:39

That analyzes the Supreme Court justices. That role was important for me in thinking about how visualizations work with stories, but also that department of interactive was fledgling when I arrived. So, I can’t say I did a lot while I was there, but I did help lay the foundation, I think for that to well into what it is now after I left. But yeah, after BPR, I did… So, not after. I joined HiGeorge last August during the pandemic. HiGeorge is a startup. They do interactive data visualizations for local newspapers, which may not have the in house team to build these very pretty interactive charts that you see in the New York Times, right? 

Jon: 00:46:23

All of these roles, they dovetail together so nicely. 

Anjali: 00:46:26

Yeah. 

Jon: 00:46:27

So, you have this interest in journalism, even before really the data science. But throughout whether we’re talking about the Berkeley Political Review or even HiGeorge, which I didn’t realize that they were involved in that space as well, you’re creating visualizations that can be used in journalism that, again, bringing you back to the idea of putting a face to data, making data accessible and understandable, coming up with clever ways. And you see this in your YouTube channel, coming up with clever ways of taking a complex concept, but breaking it down into data visualization structures that people can understand at a glance and get a lot of information from It’s so cool. Anyway, I’ll let you talk about HiGeorge. 

Anjali: 00:47:09

Thank you. Yeah, I mean, HiGeorge, as a service obviously became very important because of COVID, and people were refreshing looking at these charts and seeing, okay, how many new cases are today? How many deaths are there today? What is the hospitalization capacity? And now it’s more so what is the vaccination rates? So people got very used to reading and understanding charts, right? So it meant that I got to be a little more creative and not necessarily do basic bar charts. But yeah, so my work at HiGeorge, was we used Angular, and D3, which is a JavaScript library to build these reusable data visualization components that our data engineering team can just plug in different data sources, and the visualizations would appear. 

Anjali: 00:47:58

So, the challenge for me there was actually designing visualizations that work, no matter what the data is will look pretty no matter what the spread of data is, and what… So, you’d always have to be thinking of these [inaudible 00:48:10], okay, if the axis is this long, will it look squash or blah, blah, blah. So it was actually really interesting. I was more so doing front end web development and data visualization there. But I did have to put all my data science hat on a couple of times because I would have to think about what edge cases of the data are. And would have to create my own dummy data sets and test these visualizations and make sure they make sense and are legible. So, I think that was interesting work experience from that perspective of thinking not only is it communicable, but also doesn’t look pretty? Which is not necessarily something we always think about as data scientists. But it’s probably something we should think about. 

Jon: 00:48:51

Totally, I agree. Nice. That sounds such a useful specialization. And I’m glad that we have you to walk us through how that works. It sounds clear that this front end development aspect becomes really important when you’re doing data visualizations in production. You have been in these roles. So, the JavaScript frameworks make sense, Angular, D3. I don’t know anything else. Any other particular tools and techniques that are interesting that you use a lot in this data… these engineering roles? 

Anjali: 00:49:28

Yeah, I mean, so D3 is great because it’s super flexible. You can do anything with it, right? If you can think of an idea in your head, you can basically write using JavaScript. But I think for most of us, the visualization libraries that we end up using are Matplotlib, ggplot, seaborn, those types of things. More so than the tools I would really just emphasize learning data visualization, and what works with certain types of data and what doesn’t. I think a lot of us intuitively know that. Like the whole is, we’re like bar charts and blah, blah, blah. Numerical data is line charts. A lot of this, we intuitively know. But there are certain things you can do that just make it that much more comprehensible. Like either adding a slight annotation, adding a dashed line that works with threshold. Labeling is so important. Yeah, and there’s tons of resources. 

Anjali: 00:50:20

I really like d3graphgallery.com. Even though it’s more geared towards D3, I think the author also has one for R, and learning how to build really beautiful visualizations in R and is working on one them for Python as well. So I would keep an eye out for that. But honestly, just looking at those different tracks, and seeing what makes certain things pretty helped me a lot not just in my data visualization work, but also in my data science work, and in the reports that I end up presenting to people. 

Jon: 00:50:50

Nice. That’s really cool. I was about to ask you read my mind. I was like, do you have any resource for us to be doing this? The d3graphgallery.com. 

Anjali: 00:50:55

Yeah, I think there’s hyphens between. It’s D3-graph-gallery. You can find it. 

Jon: 00:51:06

[inaudible 00:51:06] look it up, and I’m sure Ivana, our show manager will find it on the internet and put it in the show notes. All right, so all of this journey has led you now to a point where you’re starting your first full-time job in data science. So we’re recording early July, but this episode will be out early August. So you might have started by then in your role at Thermo Fisher Scientific, which is a giant global company. And specifically, you’re going to be starting in their data science leadership development program. So tell us a bit about the company, as well as this program? 

Anjali: 00:51:41

Yeah, I mean, the company is involved in a lot of different things. I think most people are probably familiar with the name because they’ve seen it in their science equipment in high school lab or college labs. They’re involved in a lot of research pieces, but also very relevant for the time we’re in right now. They develop a lot of reagents for biologists doing research and specifically, we’re involved, I believe in the development of the Pfizer vaccine for COVID-19. So, yeah, they have their hands in a lot of different pots. But I think I’m more interested in discussing the leadership development program because I think it’s such a cool concept, and I think there should be more of them out there. It’s a rotation based program. I am spending one year in one office and one year in the next. I don’t know if I should necessarily say where I’ll be. Yeah, I don’t think I’m comfortable thing where I’ll be but one year in one office, and one year in the next. The idea is that you’re rotating between different roles as a data scientist, so you can see where your skills align, and where you’d want to be starting full-time after those [crosstalk 00:52:48]- 

Jon: 00:52:47

Nice, that is cool. 

Anjali: 00:52:50

Yeah, it’s like a nice bridge between college and full-time because you still have that mentorship, and you’re still in a learning mindset, which I think is true for most early careers, but this just makes it that much more explicit. Because there’s the idea that you’re stepping into a new role every few months, and you’re learning the ropes of what it means to be a data engineer, what it means to be data scientist, what it means to be a data analyst, and not just the specific roles, but also within, different functions. Maybe you’re on the consumer facing side, or maybe you’re more in research. Yeah, so I’m just excited all around to be experiencing so many different things in such a short timeframe. I think it will be really good. 

Jon: 00:53:32

It sounds like a great program. And no doubt, not only getting exposure to different kinds of industries like you’re saying, like commercial applications, industrial applications, or data analytics, data visualization, modeling. Not only that kind of stuff, but also the culture of the firm, and which maybe even specific people that you’re like, “Ah, man, that group was awesome. I definitely want to work with them.” And then once you actually start in the role full-time, in a specific role that whatever you decide on you’ll have a better understanding of the whole organization, and I’m sure that creates opportunities for collaboration. It just seems such a great idea of this program. 

Anjali: 00:54:17

Yeah, definitely. Yeah, I think you… Yeah, once I’m in that full-time role having that connections not only in other offices, but across the country in various places because even though I’ll only be in two offices, mainly, I think they’ve made it a point that I should be in contact with people across the country and maybe I’m here in New York, but maybe I’ll want to go into Iowa after graduating. Those are two arbitrary locations. But yeah, I don’t know. I think it’s you get to experience a lot, which is good for the age I’m in right now. [crosstalk 00:54:51]. 

Jon: 00:54:52

I’m sure everyone is queuing up to be in Iowa. That’s going to be the office that’s going to be everyone’s going to be lining up for. All right, this is super cool. I’ve loved getting to know about your journey here. I think it’s so interesting. I think that this general philosophy of putting a face to data and bringing that out to the public through your YouTube channel, your writing, which we didn’t even talk about, but related to YouTube posts, you also publish on towards data science, which is a blog. So really amazing. When you retire many decades now, what are you hoping to look back on? Do you have any idea? 

Anjali: 00:55:38

Oh, man, that is such a hard question. My dad is also someone who like me as interested in a lot of different avenues. He studied physics. He got a PhD in computer science. He was a researcher, a professor, and he eventually ended up retiring into teaching high school. And that’s something that I always saw myself doing maybe not teaching but retiring into some sort of mentorship role. Or even before that I always want to be the type of person who is teaching or sharing with people. I don’t know, maybe the YouTube channel can become that years from now once I have more things in my belt to share. But yeah, I think being able to teach and pass on what you’ve learned, and however many years that is, is really important. So maybe not something to look back on, but something to look forward to, whatever. 

Jon: 00:56:34

That was a really great answer. All right, I really only have one last question for you, Anjali, which is something I asked all the guests on the show. Do you have a book recommendation for us? 

Anjali: 00:56:44

I do. I’ve been thinking about this ever since you asked me before we started taping. So, I have been getting into fantasy. As I mentioned to you, I’ve been reading a lot. I think it’s really important to have hobbies that aren’t always geared towards being productive and wanting to be the best worker you can be even though that’s important. I think you should take steps away from that at time. And reading for me as that is how I sleep at night. So I don’t necessarily want to be thinking about work when I’m about to sleep. So, anyways, I’ve been reading a lot of fantasy. The book that I just read maybe three months ago is one of the best books that I’ve read in my life. I’ve given it to four people now. It’s called the Sword of Kaigen. 

Jon: 00:57:24

The Sword of Kaigen, how do we spell Kaigen? 

Anjali: 00:57:27

K-A-I-G-E-N. It is an Asian inspired fantasy. So, I think, especially for the child of immigrants, you’ll connect a lot to a lot of themes in it. It’s really unlike anything I’ve ever read because it’s got two protagonists, one is an 11-year-old child, and the other is his mother. So it’s a very mature story on family, and parenthood, and you just get these two wildly different perspectives. It’s very sad. I will give you a warning. It might put you out of commission for a few days where you’re like, “Wow, [inaudible 00:58:00].” But I can’t recommend it enough. It’s so good, and I think it will change a lot… I think it will just change how you see a lot of this world, honestly.
Jon: 00:58:12 It’s a great recommendation. I really appreciate that. And it’s so nice to be able to hear you speak about it so passionately. Why you’d be interested in this genre and everything. So thank you for sharing that. I’m sure we have a lot of listeners now going out and looking that book up. 

Anjali: 00:58:26

I hope so. She’s a self-published author, so she doesn’t get a lot… It’s kind of like an underground story, so I really do recommend. 

Jon: 00:58:35

Wow. It just gets cooler and cooler. So, how can people follow you? What do you recommend? So obviously, your YouTube channel, Vastava. We will definitely put that in the show notes. And that’s probably your primary way for people to stay in touch. But yeah, anything else? 

Anjali: 00:58:54

Yeah, youtube.com/vastava. Right after this is done recording I will update my LinkedIn profile, so it’s a bit more polished. You can feel free to connect with me on there, Anjali Shrivastava. I do have a Twitter. I am not as active as I would like to be on there. But you can find me on there, also, Vastava_ because the name Vastava was taken, I guess. 

Jon: 00:59:21

Brilliant. All right. I’ll be sure to include all those, the YouTube channel, LinkedIn URL, and Twitter handle. We’ll get those into the show notes. Thank you so much, Anjali. This has been so much fun. And hopefully we can have you on again sometime in the future to give us an update on how your data science career is developing. 

Anjali: 00:59:41

Sure. I’d love that. 

Jon: 00:59:43

Sounds great. All right, Anjali. Have a wonderful day and I’ll catch you soon. 

Anjali: 00:59:47

Thank you. You too. 

Jon: 00:59:54

That was so much fun filming that episode with Anjali. I hope you enjoyed it as much as I did. She’s wise well beyond for years and is clearly extremely intelligent, creative, and ambitious. No doubt, Anjali has a brilliant career ahead not only as a data scientist, but also as a broadly appealing data science communicator. In today’s episode, Anjali discussed the importance of making data analytics and data science approachable and valuable to lay people. She talked about UC Berkeley’s Data 8 foundations of data science curriculum that is available free online to anyone anywhere in the world. She talked about how we can create reusable interactive data visualization components with the Angular and D3 JavaScript libraries. And she let us know about the D3 Graph Gallery that you can refer to, to develop a better data visualization practice, whether you use D3 or some other graphics library. 

Jon: 01:00:49

As always, you can get all the show notes including the transcript for this episode, the video recording, any materials mentioned on the show, and the URL for Anjali’s YouTube channel and LinkedIn profile as well as my own social media profiles at www.superdatascience.com/493, that’s www.superdatascience.com/493. If you enjoyed this episode, I’d of course greatly appreciate it if you left a review on your favorite podcasting app, or on the SuperDataScience YouTube channel where we have a video version of this episode. To let me know your thoughts on this episode directly please do feel welcome to add me on LinkedIn or Twitter and then tag me in a post to let me know your thoughts on this episode. Your feedback is invaluable figuring out what topics we should cover next. 

Jon: 01:01:32

Since this is a free podcast if you’re looking for a free way to help me out, I’d be very grateful if you left a rating of my book, Deep Learning Illustrated on Amazon or Goodreads. Give some videos on my YouTube channel a thumbs up or subscribe to my free content rich newsletter on jonkrohn.com. To support the SuperDataScience company that kindly funds the management, editing, and production of this podcast without any annoying third party ads, you could create a free login to their learning platform at www.superdatascience.com. You could check out the 99 Days To Your First Data Science Job challenge at www.superdatascience.com/challenge or you could consider buying a usually really inexpensive Udemy course published by Ligency, an affiliate of SuperDataScience such as my Mathematical Foundations of Machine Learning course. All right, thanks to Ivana, Jaime, Mario, and JP on the SuperDataScience team for managing and producing another amazing episode today. Keep on rocking it out there folks, and I’m looking forward to enjoying another round of the SuperDataScience Podcast with you very soon.