SDS 517: Courses in Data Science and Machine Learning

This is episode number 517 with Sadie St. Lawrence, data science instructor at the University of California Davis, and founder and CEO of Women in Data. 

Jon: 00:14

Welcome to the SuperDataScience podcast. My name is Jon Krohn, chief data scientist and best-selling 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: 00:44

Welcome back to the SuperDataScience podcast. Today’s guest is the fun and immaculately articulate Sadie St. Lawrence. Sadie is an instructor at the University of California Davis, whose data science course is one of the most popular on the Coursera platform with over 300,000 students all time. Sadie’s also the founder and CEO of Women in Data, an organization that increases diversity in data careers, and that has grown to become a community of over 20,000 people across 45 chapters and 15 countries. She also holds a master’s in analytics from Villanova University. In today’s episode, Sadie digs into the content of her existing data science course, as well as the curriculum of her forthcoming UC Davis machine learning certificate, which will be available on Coursera as well. She talks about the mission, impact, and vision of her inspiring Women in Data organization, her path into data science from music performance, and non-fungible tokens, and the future of technology. Today’s episode has a lot of laughs, and so we’ll appeal to anyone who likes to enjoy themselves while learning at a relatively high level about the leading tools and approaches in the field of data science and machine learning today. All right. You’re ready? Let’s do it. 

Jon: 02:18

Sadie, welcome to the podcast. I’m so happy to have you here. Where in the world are you calling in from? 

Sadie: 02:26

I’m calling in from Sacramento, California. 

Jon: 02:29

Ah, yeah, Sacramento, California, also the hometown of Harpreet Sahota, which is how we kind of know each other. It’s how we were introduced and how I landed you as a guest on this show. Harpreet Sahota, amongst the incredible large number of data science community, things that he does, one of those things is the Artists of Data Science podcast, which is great. In mid-September, the two of you did a live episode on Instagram, right? 

Sadie: 03:04

Yes, we did. Yeah. I had the pleasure of being on his podcast about a year ago and we’ve stayed in touch, I think, just from the books we read, and what’s going on, and the fact that we share hometown. It was so fun to do an Insta live because we noticed that our communities both have similar questions and it’s just more fun when you get to do things with friends. Yeah, it was a great time. People tuned in from all over the world, got to ask their questions about how to get into data science, how to stay relevant in the field. I think we dropped some wisdom on them, hopefully. 

Jon: 03:38

Yeah, I love it. I saw that you were doing that, that you were doing this live on Instagram podcast with Harpreet, and you’d come up in my LinkedIn world a number of times. I’d seen your name, I’d looked at your profile, and I was like, wow, Sadie looks super interesting, but I didn’t know that I knew anyone that could kind of make an introduction. Then when I saw that that episode was happening, I immediately texted Harpreet and was like, “You think Sadie would want to be on the show?” And here you are, it happened. Sadie, you have done a lot on Coursera. It also seems to be kind of through the University of California Davis. So, I’m going to give some kind of summary stats on how incredible your data science course has been received on Coursera. It’s had over 300,000 students, which is crazy. That makes it one of the most popular data science courses in Coursera. But on top of that, it’s also one of the most highly rated. Interestingly, you were the first female data science instructor in the platform. Some interesting stats there. Tell us a bit about the course. I’m interested in actually, first of all, how it’s both UC Davis, but also Coursera, what’s the link there? 

Sadie: 05:03

Yeah. Great question. The course is really designed for people with no coding background and an introduction to SQL. I was excited to teach SQL because it’s so fundamental when you’re beginning your data science journey. I think it’s the first thing anyone should learn because SQL really is about how to read and write data, right? So, if you want to get into a data career, reading and writing is the first thing you learn. When I got approached by UC Davis to teach the course, I was a little hesitant at first just with what was going on in my schedule, but I actually started my data science journey on Coursera, and so when that opportunity came up, it was just like, had to say yes, as a way to give back to the community. How UC Davis is involved is in the fact that, up until a few years ago, Coursera not put any classes on their platform that weren’t associated with the university. 

Sadie: 06:02

Coursera, at its core, is really a platform. Where they’re saying, hey, we’re hosting the tool and bringing people together, but it was up to universities to provide that content. Unlike Udemy, you couldn’t be an individual and just say, hey, I have this awesome class I want to share it. You had to be associated with a university. Now they may be changing a little bit. I know that Google has a certificate up there and a few others are coming in, but that was originally how it started, was purely with universities. 

Jon: 06:35

Nice. That does ring correct. I must admit, I don’t think I’ve done a Coursera course before, but I had this impression that it does … It seems to me like a lot of the Coursera content I see out there is from universities, or as you say, highly regarded institutions like Google. I guess, so that gives it this quality edge where they can say, we curate this content. We make sure that everything in here is from a top university or a top institution. Unlike Udemy, which does have some absolutely amazing content, including from SuperDataScience in there, some of the best selling data science courses are in there, but in Udemy, anybody can publish. There’s a review before you get to publish in there, but it isn’t so much a review to be like, yeah, this is a UC Davis level quality course. It’s like, okay, this isn’t bad. So, cool. Coursera, yeah, great platform for learning in, and it makes a lot of sense to me that SQL is something that you’re so passionate about teaching because it is the most widely used data science tool, according to data science surveys, which I’m sure you already knew. 

Sadie: 07:56

Yes. Thankfully, at the time I was a practicing data scientist. When I got approached to teach a technology course, I was very cautious about which language and tool I would choose, because I knew how rapidly changing the industry was, but sequels like the OG of data science, the OG of like coding anything. I’m like, this is not going away anytime soon, and I think that really helps with like the popularity of it as well. It hasn’t changed and the course has been on the platform for about five years. 

Jon: 08:33

Yeah, it’s definitely a critical tool. I guess the reason why it doesn’t change, we get different flavors, right? SQL in general, structured query language, is this way of writing data queries of structuring your queries in a specific way, you know it much better than me, but we can say things like select all of the rows in this particular database that meet these particular criteria, and then boom, it gives you those results back, and it does lots of clever things in the backend to allow that data to be pulled in an efficient way. So, it doesn’t do it in a haphazard way. So, from that kind of like root idea of having this efficient, structured querying language, there’s so many different flavors out there. I mean, there must be dozens and dozens, but like MySQL is a popular opensource one, but there’s like … Yeah. 

Sadie: 09:29

SQLite and they go on and on. Yeah, but I mean, I’m so thankful we do have a standardization for it, because could you imagine if every database that we had to read from, we had to learn a new language for, and then it’s already hard enough to get databases to talk to each other and combine data. If there wasn’t a set language for it, I think our lives would be even more miserable as cleaning data, but also, we wouldn’t be able to have the impact with data that we do. I really hope to see the same thing in algorithms, because I think once we get them more standardized and how we’re building them and using them, we also have an opportunity to then add in how we evaluate them and the ethics of them. I think just a little bit more standardization across the board in data science could be really helpful for us as well. 

Jon: 10:20

Yeah. It seems like such an obvious idea that I’m sure those things will happen more and more. Some particular algorithm libraries like scikit-learn are kind of defacto standards, but then there are some people who only do things in R. So, it’s interesting where we are kind of lucky that with pulling data from databases, everybody’s kind of said, okay, the SQL way of doing it is fine, maybe we’ll have our own little flavors, but luckily, unlike with when we’re actually deploying an algorithm, and there’s so many different languages in libraries we could choose for getting your data out of a database, you can learn SQL once and it’s going to work for your whole career. That is really convenient.

 

Sadie: 11:03

Yeah. I think there are people still holding on to NoSQL a little bit, but we don’t hear about them as much, so I think you’re still [crosstalk 00:11:11]. 

Jon: 11:12

Well, that is a really good point. I mean, so I use NoSQL every day and we use NoSQL databases. You’re right. There is that main … Yeah, it depends on how you store your data and what you want to do with it. The convenience of SQL is that every single table has exact same columns and the exact same rows. So, with NoSQL, the thing that makes it more flexible, but often then a lot more challenging to work with, is that you just have this big list of documents in a given database, and you can have any fields you want in any single document. So, yeah, you’re right. I can’t believe I didn’t even think to mention NoSQL. Ooh, embarrassing. 

Sadie: 11:57

See, I had to mention them. I knew they’d raised their hand eventually because the NoSQL people can’t stay quiet. 

Jon: 12:02

Oh no. And it is that adaptability, there’s certainly circumstances. For us, with developing our product, you might, over time, want to have the flexibility to say, okay, in the future, we’re going to want to have different fields. We don’t want to have to constrain ourselves to specific columns. You can have these changes in time in the product and the way the things get logged. But then ultimately, very often, if you want to be able to make quick calls to those data, if you are using NoSQL for that flexibility, then you’re going to have all of this complexity that could be unknown in the way that all of these different documents in the NoSQL database are structured. You kind of have to have a lot of documentation as to like, okay, in 2017, this particular field was called this, but if you want it from 2018 onward, it’s called this other thing. We used to do it in metric, but now it’s imperial, and were going to have to do this conversion. So, you kind of have to have that know-how. Ultimately, for us, when we want to be able to do analytics quickly, we have scripts that pull information out of the NoSQL databases every night at midnight and put it into SQL databases so that we can very quickly access it. I don’t know. Maybe that gives our listeners some sense of why you might want one or the other, but SQL, it is the most popular. Speaking of popularity, what do you think it was about your course that led to it being so popular? Do you have a unique presentation style or just a clear way of presenting things? How did this take off and be so big? 

Sadie: 13:49

Yeah, I mean, I really came in to the course with the mindset, if it will just help one person, I will be happy. I think that is helpful going into somebody, is having a really good target of your audience and having one person in mind with who you’re designing it for. As you can see, obviously more than one person has taken the course, so I was able to scale up, which is great, but I really think it comes down to- 

Jon: 14:13

300,000 times more people than you anticipated, and that’s just so far. Yeah. 

Sadie: 14:19

But I really think it comes from my background in teaching music. We chatted a little bit before that we both have a piano background and through college I taught piano lessons. Through that experience, I realized there was a much different skill from teaching versus doing. Just because I could use SQL and was good at it, didn’t mean that I was able to teach it. I think that’s where we see a lot of technical classes go awry. They get the expert in the field who’s good at doing, but teaching is a completely different skill. I found that out through teaching piano lessons where I realized that I would need to know something so well that I could explain it 10 to 15 different ways so that it would resonate with that person. When going into this class, I had a very specific target audience in mind and made sure that how I was talking about the language was in regards to what would resonate with that person versus how I just talk about SQL and use it in my day-to-day life. They’re two very different things. I think that was really important. I came from a non-technical background and a lot of people who are taking this class are coming from a non-technical background, so I will make it as simple as possible and give them the foundation steps, which really, I think helps people be successful in the beginning and then helps them keep going along the way to be able to finish the class. 

Jon: 15:50

Nice. That was a wonderful explanation, and it makes so much sense. I realized, after I asked the question, that I was really putting you on the spot with, why is your coerce so popular? But yeah, that’s a great answer. The piano teaching makes perfect sense. Yeah, you have fundamental concepts that could be quite complicated. Then you have lots of different kinds of piano students come in who need to get that same piece of information told to them in a totally different way relative to other students. That makes so much sense to me as the kind of secret behind why your teaching methodology is so effective. Speaking of which, that course that is available in Coursera already, is not the end of the road for your teaching. You’re currently designing a machine learning certificate for UC Davis. It sounds like this is going to be available around spring 2022. What’s in this course? What’s it all about Sadie? 

Sadie: 16:55

Yeah. I’m really excited about this because it is actually four courses and it’s really targeted from an applied machine learning aspects. Just as I mentioned with SQL having a really specific target audience, for this course, we have a really specific target audience as well. It’s someone coming from maybe an engineering background, someone coming from an analyst background but are looking to add in those machine learning skillsets to deepen their skillset and then go and get a job in it. This will be an applied machine learning certificate, and I’m really heavily focused on the project standpoint of it. The goal with this is that people can take these classes, the four classes, get the certificate, and then go use it right away in their job. We’ll go through an introduction to machine learning, go through supervised, unsupervised, deep learning where we’ll be exploring NLP. We’ll also be doing some computer vision. Then, what I’m excited for as well is taking the dive into cloud computing, technology and architecture and how that plays such a vital role in machine learning. From the end of this project, people will have four core projects that they can share and talk about what they’ve done. I’m just excited to see what students come up with in terms of the projects that they do and the careers that they go on and have after this certificate. 

Jon: 18:33

Yeah. Eliminating unnecessary distractions is one of the central principles of my lifestyle. As such, I only subscribed 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, 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. 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. 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. 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. Now, let’s return to our amazing episode. 

Jon: 20:35

That sounds hugely valuable. That certificate, the four courses in it. If I got this right, I was trying to jot it down as he went through it and I was kind of parsing, I was guessing. Okay, I think this is a course. I think this is the course. It sounds like the first one is an intro to machine learning and then the second one is on supervised and unsupervised learning. The third one is deep learning and then the fourth one was like, cloud-based … Explain that one to me again. It was like using, so you’re kind of leveraging tools, I guess, maybe things like AWS SageMaker or these kinds of these built-in cloud provider machine learning tools that allow us to really quickly maximize the way that we’re applying machine learning. Our hyper parameters searches are automated. Things are deployed in an efficient way automatically. I’m guessing that’s what that fourth course is about? 

Sadie: 21:28

Yeah, exactly. The names of the course aren’t quite ready. You did a pretty good job, three out of four [crosstalk 00:21:34]. I’m already grading you here now. The first one is the intro to machine learning, and then we’ll go over the supervised and unsupervised learning, and then there’s the applied machine learning. This will be also looking at not only, how do you do the algorithms, but how do you deliver this to production? Who are the other team players that you’ll be working with? Because I truly believe that data science and machine learning is a team sport. Then the fourth class will be the deep learning class. 

Jon: 22:08

Okay. I see. I see. I see. Very cool. I love that curriculum. If I was getting started in machine learning, it sounds like you’ve done your homework because that is perfect. I like that you do have this focus on the applied ML in that third course, because I think that, that’s where a lot of courses are missing material. A lot of courses out there, we’ll do the intro to machine learning. Okay, and then, once we’ve done the intro, supervise approaches where we’re predicting some particular class with our model or predicting some particular number with our model, as well as unsupervised learning, where we’re just trying to learn some structure about some unlabeled data. That makes perfect sense. Then also teaching the deep learning. It can be a relatively advanced topic. Really amazing applications. Like you mentioned, computer vision, natural language processing. A lot of ML courses aren’t complete without relatively advanced topic, but it’s too easy to be missing that third one, the third of the four courses there on applied machine learning. You’re absolutely right that creating a model doesn’t happen in a vacuum. It needs to be delivered into a system where people can use it, and getting that right. Is, is key to the success of your machine learning application. So, very cool. I don’t know if you have anything more you want to say about that, but I think this is awesome. 

Sadie: 23:38

Oh, thank you. Yeah. I’m always looking for peer reviews, so appreciate the quick peer review. Yeah, I just say one thing on the applied side, that’s probably one of the portions I most excited about. Because when looking at the courses, I went and evaluated, what do I wish I would have known, it wasn’t taught in my master’s program? Most education is good on the theory, the tools languages, but they don’t tell you like, who do you need to go talk to in the business to get something done? How do you work with data engineers to implement your pipelines into production? All those kind of secret sauce, things that you learn just on the job. I really want to give people a headstart, in that, in terms of using it for their career and using it right away in an applicable way to work. 

Jon: 24:27

I think it’s great. I know how important it is, because from my own experience teaching my deep learning curriculum, that’s one of the most frequently asked questions I get is, how can we … Okay, this is great, we’ve learned this, but do you have any content on how this can be engineered and how we can make use of it in production? I’ve historically just shied away from it. I’m like, there’s so many different ways you could do it. It depends on your particular circumstances and you’re just going to have to kind of look it up on your own and figure it out. So, good on you for like tackling it and doing it, which I’m … Just because I shy away from it, that I now feel like that definitely isn’t the right thing to do and you’re doing the right thing here. Maybe in the future, when people will ask me that question, I can say, check out course three of Sadie’s machine learning certificate. Sounds like the perfect thing that you need. Cool. You’ve had this tremendous success with your teaching in the past. I have no doubt that this course is also going to be massive, but that isn’t the only thing you do. You are also the founder and CEO of Women in Data. It looks like you founded it in 2015, so six years ago and it’s grown quite a bit. You now have over 45 chapters in 15 countries, a community of over 20,000 people. Tell us about what the Women in Sata organization is. Why did you found it? Maybe let’s just start with those two questions. 

Sadie: 26:03

Okay. Sounds good. Yeah, so Women in Data’s mission’s pretty simple. One, because I’m not a marketer, I’m a data scientist at heart, so the name is very self-explanatory. We are Women in Data. Our mission follows the same suit, right? Our mission is to increase diversity in data careers, and we do this through our three pillars of awareness education, and then advancement, which also goes in the linear line as well. You see, I love geometry, I love things crystal clear and very much the core of Women in Data. But yeah, essentially we measure two things, which is, how many people are we getting into data careers and how many are we helping stay and advance in their career? If you think about how we increase diversity, we have to fill the pipeline and we also have to make sure people aren’t falling out at the end of the pipeline as well. So, we have a variety of programs in our pillars. We do weekly global webinars. We have networking events with our chapters and globally. We also have a residency program, which is a kind of internship for career transitioners. We do study groups and certifications through data camp. We have mentorship, life coaching, so lots of different programs to help our members throughout their career journey. 

Jon: 27:25

Yeah, that sounds amazingly comprehensive. I was too slow in writing down the three pillars. Remind me what they are again. 

Sadie: 27:31

Yeah. So, it’s awareness, education, and then advancement. If you’re brand new to the field, we look at it as like, okay, let’s just open your awareness to, what are the jobs in the data science family? Because there’s lots of different careers, and that’s why I call them data careers. It’s not just data science. Then once you’re aware, you probably need a little education, and then once you’re educated like, how do we advance you in what you’re already doing? 

Jon: 28:00

Wow. Yeah. It sounds like from all of the programs that you covered, in response to my preceding question, it’s so comprehensive. I mean, that really does sound like you’re nailing down all three of those pillars, awareness, education and advancement. Yeah, so cool that you’re monitoring the impact that you’re having, quantitatively tracking how many people you’re impacting, and then also tracking that they’re able to, as you say, stay in the pipeline to succeed at the other end of it with I guess, the advancement pillar. That is super cool. I don’t suppose, Sadie, you have … Actually, I was going to ask you for a couple of interesting case studies, but before I even do that, how can people get involved in women and data themselves? How can they, I guess, you have a website and they can sign up or is there an application process? How do you get into Women in Data? 

Sadie: 28:54

Yeah. You can just go to womenindata.org and there’s a tab there to become a member. So, you can sign up to become a member. You can also become like a partner. If you are looking to hire more women, there’s ways to partner with us there, you can also speak at some of our events. So all of the, that information is on the website. 

Jon: 29:17

Nice. Cool. All right. And then, so I don’t suppose, going back to the question I was just about, circling back, do you have some interesting case studies of impact that you’ve had? 

Sadie: 29:31

Yeah. I’m so glad you asked it for case studies, because a lot of times what gets reported is what I call vanity metrics and I know we chatted a little bit about this before. 

Jon: 29:43

[crosstalk 00:29:43] any metrics. Personally, people go and look at my LinkedIn page. At a glance, you can see from the quantities, somebody is having an impact. I don’t know, but yes, it isn’t very warm and fuzzy feeling to look at these vanity metrics, I agree. Yeah, case studies are great too. 

Sadie: 30:03

But yeah, from a social impact organization, I really love the case studies because it gets to the heart and core of the stories behind it. One of the studies I love is from a participant who joined our residency program. As I mentioned, our residency program is an eight week program, that we take a team of 10 people through a data science project for a company or a local government, nonprofit from beginning to completion. This is a way for them to get work experience, because what we found through our community is a lot of people are going and getting education either through Udemy, Coursera different universities, but then when they go to apply, they don’t have the job experience. So, this was a way to help support that. One of the participants through that program completed the program and then was able to get her job in, her first job, in data science. She just called me crying and thanking me because she was able to get a job that paid her over $100,000, which allowed her to leave an abusive relationship that she was in and now could support herself and her family. I was blown away by that impact. It really just goes to show like, when you support women, they just give so much back to their communities and their families as well. Not only did it change this woman’s life, but also changed the prospects for her kids and the safety of the environment that they were living in as well. 

Jon: 31:39

Wow. That is beautiful. Such a great story. Sadie, what’s your vision? You have these clear pillars, and it sounds like you’ve already covered a lot of the support that people need across all three pillars, awareness, education, and advancement. You already have quite a bit of scope, so 15 countries, 45 chapters, 30,000 people. 20,000 people. 20,000.I don’t want to overstate the vanity metrics. Obviously continuing to grow the impact, to scale the impact, that’s obviously going to be one of your objectives, but is there anything else? Is there anything else as part of your vision? Yeah. 

Sadie: 32:25

Yeah. Right now I’m really excited because we’re launching a new program called the Path Forward, and it’s for women who were affected by COVID-19 and dropped out of the workforce. We’re giving them kind of the full suite of our programs, but in a synchronized fashion to get them back in the workforce. I’m really excited for this program because it is very timely and relevant for what’s happening right now, but also just with the prospects of giving them a better career. I think the long-term vision though, is how do we go from just supporting women in getting into a data career and staying in there, to helping all individuals become data literate. I really look at data science as a tool set. Right now I’m not working so much as a practicing data science, but I use data science all the time to run my business, Women in Data. I think, as we move into more of a digital economy, it’s highly important that all of us have those data literacy skills, not only just to read media, to understand news, but to be a participant in the economy. That’s what I would really love to see us move into in the future is like, how do we empower people to become data literate and use that to make better decisions in every area of their life, whether that be business or personal. 

Jon: 33:58

I totally agree. That’s a beautiful vision. Yeah, so whether it’s as a consumer or as a business person, we already have so much data available to us. And if you can’t write a SQL query, you can’t even really get started. Yeah. There’s such an opportunity regardless of whether you’re a professional data scientist or not, to have great data literacy. Yeah, as you say, even just with reading the news, critically thinking about claims, data visualizations is a hugely valuable skill for anyone. Then of course, being able to apply it yourself to make better consumer decisions, to be able to understand how things are growing in your own business, keeping track of where things are doing well, where things aren’t going well, where you need to focus your attentions. Your attention. I don’t know I’ve pluralized that. Yeah, so that’s a beautiful vision. I love it. You mentioned earlier on in the episode, I thought this was really funny, and I almost derailed the podcast back then to bring this up, is that you said, well, I come from a non-technical background, and I was thinking … You said that in the context of how that, it helped you be such a great SQL instructor. I thought that was interesting, because at some point, everyone [crosstalk 00:35:32] background. But I think I know you meant with having the music teacher experience already. Let’s talk a little bit about your path into data science because even before, I think you made this SQL Coursera course, you already, at that point, I think you had your master’s in analytics from Villanova, right? 

Sadie: 36:00

Yes.

Jon: 36:00

Your technical background was blossoming already at that point in your career. You had the music teaching background, which has been hugely useful for being able to communicate and teach effectively. But then you transitioned into doing a psychology degree at California State University, Sacramento before doing the Villanova masters in analytics. I don’t know if you’d like to share with us a little bit that transition. Why the shift from music to the sciences and then from the sciences to data analytics in particular? 

Sadie: 36:40

Yeah. Happy to share. I love your comment on the non-technical background because … You’re actually right. It’s not that babies are born just technical or not. I think it comes from the idea of like the Bill Gateses of the world where they were coding at like 10, 15 and onward. But I think the majority of people in this industry had a previous life and a career before they transitioned into data science. But my journey into data science, as you mentioned, it started in piano performance. That really came from the fact that I didn’t know anything else going into school, so I was homeschooled. I grew up on a farm in Iowa with six siblings, lots of different pets and animals and things like that around 

Jon: 37:32

Wow, and a bunch of pianos.

 

Sadie: 37:35

Yeah, obviously a lot of time to practice the piano. It was just kind of obvious at that point of like, okay, well, this is what I’m good at so this is what I’ll do. I got into school, and one, I realized I was smart, so I never had a grade in my life, I had never taken a test until I got to college. I went into school. I was like, oh my God, I’m actually smart. This is kind of [crosstalk 00:38:01]. I can do something with this. That was a fun surprise. Then yeah, just started taking GE classes and really fell in love with science and things. And honestly, really, I would say just the scientific method. I think the scientific method is so beautiful, like the SQL language, like it’s a standard thing we have to analyze what is true and ask new questions of things. My transition into psychology really came from just like falling in love with science. I felt that psychology was a good bridge between music and not going into physics or any type of that. That actually may play now that I think of it, but it felt like a good bridge at the time. I was looking at studying, I was really interested in studying music and how our brain interprets that. I tried to get in a bunch of labs to start studying that. No one wanted me at first. I started out in a fish lab studying why fish eat their young, which is a little depressing subject. Yeah. You gotta really start on the bottom to … 

Jon: 39:22

I guess you have to create circumstances where they’re likely to eat their young.

Sadie: 39:27

Yes. 

Jon: 39:27

In a lab, wow. 

Sadie: 39:27

It’s called filial cannibalism. Then I got to upgrade a little bit to emotional learning and memory. Then got to work in that loud for a little while. Again, I really love the science aspect of it, but I hated working with animals. One day it was a Friday, I went to euthanize a rat. It looked me in the eyes. We had like this moment of like, why are you doing this to me? And I was like, yeah, why am I doing this to you? I questioned everything and found data science on a Google search, and I was like, yep, this is how I can keep the science aspect of it without doing harm and without having to watch bats every day run around in the hamster wheel. 

Jon: 40:16

It’s uncanny, Sadie, the parallels in your career and mine. Viewers wouldn’t know this because I’ve never mentioned this on air, but at the beginning of my undergrad, I also had the opportunity to major in music, in my case was, it was vocal performance. I too was smitten with the scientific method and I loved this idea of having a way of learning the truth about the world, this methodology. So, I majored in biology and psychology with a minor in chemistry. I was in similar kinds of labs. I was in a lab working on rats, doing brain surgeries, because I had this idea at that time, I was like, oh, I think maybe I’ll go to med school, and working with the rats and euthanizing the rats at the end of my experiment was such a horribly traumatizing experience. I was like, I’m not going to do that again. Then when I got into a neuroscience PhD, at the beginning of the neuroscience PhD, there was this, it wasn’t a mandatory, but every single … So, there were 20 of us that went into my PhD program, and every single other person did this animal handling course. I was like, I’m not going to do this because I’ve done it before and I am not going to work on animals in the future. At that point, while most of the people in my neuroscience degree did work with animals or with at least tissues, and I think that, that’s … Tissues don’t have feelings, so that’s okay. But it was obvious to me at that point that studying data science tools, at the time, the term data science didn’t exist because I’m dating myself. 

Jon: 42:05

But you were at these transition points later than me. Data science didn’t exist, but it was obvious to me, okay, I’m working in neuroscience. We have huge amounts of data from brain imaging, from whole genome studies, and so I can learn to specialize in sitting at a computer and teach myself SQL queries, teach myself R teach myself Python, teach myself how to train machine learning models, how to create new machine learning ideas, and so that was my focus. Yeah, anyway, I’m in a way replaying back to you your things, but also telling you my story and telling the audience a story that they didn’t have before. I totally understand every step of your journey. It’s so exciting to see you making the impact that you’re making. 

Sadie: 42:55

That’s so exciting. I think you’re the first person I’ve ever met who had a traumatizing grad experience as well, and enough to make this transition into data science. 

Jon: 43:09

I mean, the way that we euthanize them … You develop this experience. You do the experiments with the rats over weeks or months, and so you really do know them. Then, at the end of the experiment, you have these warm moving rats that you’ve gotten to know over the preceding months. In our case, we put them all in a plexiglass box that you close and then you pump carbon dioxide into it. And you wait until the rats stop moving and their eyes turn from pink to blue. Once all of their eyes have turned blue, then you know they’re done and then you open the box and you take their cold lifeless bodies, but they’re so warm. Actually, they’re not even cold, they’re warm still and you put them in a garbage bag and they get incinerated. It’s crazy. I mean, a lot of people do it and a huge amount of valuable research comes out of it. I don’t want to devalue huge breakthroughs and things that we can do for people in Alzheimer’s research, Parkinson’s research, countless neuroscience and other disorders that humans face. I think it’s a hugely valuable thing, but that particular piece of work is not something I was … I didn’t want that to be my life. 

Sadie: 44:24

No, I completely understand. We did a little different method, and I’m not going to traumatize people on the podcast, but talking about all the ways you could kill a rodent, but yeah, at the time, I was also learning how rodents laugh. I was studying emotional learning and memory, like emotion. I’m studying their emotion. How can you look at something and you’re studying emotion and then go and understand that they have emotions and then still do that to them? Definitely a moral dilemma in that case. 

Jon: 45:01

Totally. Yeah. We had debates in … Our faculty at the university I was at, at that time, we had debates where literally every faculty member, and because I was in a lab, I was invited as well. Some people would really take that, it’s like, look, we are studying this, exactly like you’re saying, we’re studying these as a model of how our minds are. So, therefore, and from the research that we are doing, it is clear that animals have emotions and consciousness that is very similar to what we have as humans. Then, isn’t there, as you say, this irony that we’re then putting them in these circumstances. Anyway, tough moral questions there that I don’t think we’re going to end up resolving on this podcast today, Sadie because huge value comes from the work. But yeah, anyway, we shouldn’t take that kind of work lightly. Anyway, super cool to hear that about your past. Sadie, do you have a book recommendation for us? In your wisdom, all the different kind of lives you’ve lived in a way, all the different paths that you’ve taken, I’m sure you’ve come across some really interesting literature. You got a book recommendation for us?

 

Sadie: 46:29

I do. This one comes because I love thinking about the future and love learning new technologies. And it has been a huge inspiration in my life. Right now I’m really into NFTs and NFTR, and VR, and AR as well. 

Jon: 46:47

Oh, whoa. These acronyms are coming out. I barely know what these acronyms are. Okay. So, VR, virtual reality. NFTs, non-fungible tokens. Right? 

Sadie: 47:00

Yes. 

Jon: 47:01

Maybe you’ll need to tell me a little bit about that and probably at least a few of our audience members. This comes from the cryptocurrency world, but so the idea of … I guess one Bitcoin is a token, but Bitcoins are fungible just like barrels of oil, to a certain standard, where you can consider one barrel of oil that is a certain standard to be fungible. You can replace them, or dollar bills are fungible. Non-fungible tokens are tokens that are unique. They could be a piece of art, I guess. I’ve seen that’s come up. 

Sadie: 47:40

That’s probably the most famous one is people who sold a digital piece of parts as a JPEG, essentially for $69 million through Christie’s Auction. I think, to date, there’s been about $6 billion sold in art through NFT. So, this market is ginormous, a ton is happening, but it doesn’t have to just be art. Essentially, it’s just a record of something someone owns on the blockchain. It allows you to validate authenticity of it. I see this changing the fashion industry. I mean, you want to make sure that those Jordans are real, right? Eventually, those will have an NFT tied to them and we can see who originally owned it and who it’s been passed on to. But I got into this space because I got into painting abstract art during the pandemic when we all had a lot of extra time on our hands. So, went a little crazy down that avenue, found out about NFTs and people selling their art on there, and I was like, oh my God, this combines multiple loves together, which is technology and art. And now it’s just been a rabbit hole of where I spend my time. 

Jon: 49:00

Cool. And then I cut you off when you were going to give us the book recommendation. Now that we know what NFTs are, tell us about, you were talking about NFTs, VR. You’re probably getting to some really cool technology book. 

Sadie: 49:12

Yes. So, it’s actually not a technology book. It’s a science fiction book, Ready Player One, and if you’re really lazy, you can watch the movie. I think that everybody should watch this now, because if you really want to understand where society is going and what the future looks like, I think that it’s the perfect depiction of that. I give artists a lot of credit in technology because they’re usually the ones who come up with the ideas and lead the way. Facebook actually just came out this year and said, they’re building the Metaverse, which is like the representation in the book. Usually, I’m a big self-help book, but right now what’s relevant to me is that everybody knows about the Metaverse and NFTs, blockchain, and either reads or watches Ready Player One so they can understand what the future looks like. 

Jon: 50:06

Awesome book recommendation. Sadie, you have had such valuable guidance for us in this episode related to SQL, the machine learning certificate that you’re developing, Women in Data, your path into data science, and now, Ready Player One, and NFTs, and so on. Clearly, there’s a lot to get from kind of following you continuously. How should people do that? 

Sadie: 50:35

Yeah, so the platforms I’m most active on are probably Instagram, Linkedin. You can find me very easily. My name’s quite unique such that Sadie St. Lawrence is all my social handles, and same thing for my website. I’m very lucky my parents named me a unique name. 

Jon: 50:57

Nice. Yeah, it’s true. Being Jon is not super convenient. A lot of guests, by far, the most common social media for guests of the show is LinkedIn, and it’s my most active social medium. And I must say I neglect Instagram. I have a private account with like a hundred friends that I’m really close to. Literally, I have a to-do that comes up every couple of weekends in my to-do list that’s like, you should post something on Instagram. Otherwise, I never go in. So, I download the app, I post my one thing. I like some things that come up in my feed and then I delete the app off my phone. I am not making the best use of Instagram. For all those SuperDataScience listeners out there, any of those that are really active on Instagram, I think you’re going to … You really corner the market as a SuperDataScience guest [crosstalk 00:51:50] to you, so that’s really cool. It’s awesome that you do things like live episodes, live podcast episodes, like the one you did with Harpreet on Instagram. So, it sounds like I’m probably missing out on something here and I should be taking advantage of it. But very cool. I don’t know if there’s anything else about Instagram or the Instagram community that you want to share with us. 

Sadie: 52:13

Yeah, I think I just love it because it’s … I try and share a little bit more real life behind the scenes type of stuff. I know LinkedIn kind of came out with stories, and that was their impression, but then, if you’ve seen the article, they said, yeah, we’re getting rid of stories. So, people just still today, have a really hard time, I think on LinkedIn, not having that business professional space. If you want to just hang out with my … Meet my dog, see me really frustrated with some data science problems I’m working, and then get to do some Instagram lives every once in a while and chat with people, Instagram is the place for that, but I feel like you need to come on an Instagram live now so that we can bump up those followers and get you to use that more than once a week. 

Jon: 53:05

Yeah. I mean, that sounds like a great idea. I’m down. I will do it. I got to see what it’s all about. 

Sadie: 53:12

Cool. We’ll [crosstalk 00:53:13]. 

Jon: 53:14

Sounds good. All right, thank you so much for being on the program. This has been such a fun episode. You’re a wonderful guest, and hopefully we’ll have you on again sometime soon. 

Sadie: 53:24

Thanks so much. 

Jon: 53:31

Sadie was an exemplary podcast guest. We didn’t have a single retake. She’s remarkably clear and succinct with all of her communication, and in case it wasn’t obvious, I thoroughly enjoyed filming this episode with her. I hope you enjoyed it as much as I did. In the episode, Sadie filled us in on the wide utility of SQL and how it contrasts with NoSQL databases, the practical project oriented focus of her forthcoming machine learning certificate, which will consist of four courses, intro to machine learning, supervised and unsupervised learning, applied machine learning, and deep learning. She also talked about the broad ranging initiatives of her Women in Data organization across its three pillars of awareness, education and advancement, and she talked about the value of data literacy for consumers, business people and technical experts alike now, and even more so in the future. As always, you can get all the show notes, including the transcript for this episode, the video recording, any materials mentioned on the show, the URLs for Sadie’s LinkedIn and Instagram profiles, as well as my own social media profiles at www.superdatascience.com/517. That’s www.superdatascience.com/517. 

Jon: 54:54

If you enjoyed this episode, I’d greatly appreciate it if you left a review on your favorite podcasting app or on the SuperDataScience YouTube channel. I also encourage you to let me know your thoughts on this episode directly by adding me on LinkedIn or Twitter, and then tagging me in a post about it. Your feedback is invaluable for helping us shape future episodes of the show. All right, thanks to Ivana, Jaime, Mario, and JP on the SuperDataScience team for managing and producing another fun episode for us today. Keep on rocking it out there, folks, and I’m looking forward to another round of the SuperDataScience podcast with you very soon.