Kirill Eremenko: 00:00:00
This is episode number 389, with Aspiring Data Scientist, Josh Hortaleza.
Kirill Eremenko: 00:00:12
Welcome to the SuperDataScience Podcast. My name is Kirill Eremenko, Data Science Coach, and Lifestyle Entrepreneur. And each week, we bring you inspiring people and ideas to help you build your successful career in data science. Thanks for being here today, and now let’s make the complex simple.
Kirill Eremenko: 00:00:37
Welcome back to the SuperDataScience Podcast, everybody. Super pumped for you to check out this episode. This is going to be epic. The level of energy that Josh brought today is incredible. We just finished our call. I’m still super pumped, super energized. This is going to be one of the top episodes on this podcast.
Kirill Eremenko: 00:01:04
Josh is an aspiring data scientist. He has a fantastic approach to his career. So, this podcast is going to be especially valuable to you if you are breaking into the field of data science. Whether you are a student who just finished your degree and you’re looking to get a job in data science and build a career, or you’re transitioning from another area, from another industry, from another career into data science, you’re going to find this extremely valuable.
Kirill Eremenko: 00:01:34
So, the things we talked about are how to look at internships in data science. The way Josh looks at them and how he uses them and he’s on his fifth internship right now, and they all last four months. You’ll find out exactly why this is by design in this podcast.
Kirill Eremenko: 00:01:52
Also we talk about applications of different areas of data science in different industries, for instance in LP and banking, CNN’s commercial neural networks and insurance, and perhaps a few other examples. But mostly, it was about the career, how he structures his journey, how he chooses the companies he wants to work with and where to get these internships. What role luck plays in his journey, and how he minimizes luck or actually in reality how he maximizes his luck, how he uses networking.
Kirill Eremenko: 00:02:26
We talked a lot about networking. So, we get some very valuable tips about how to genuinely network with people, how to be genuinely interested in what they do. So, you’ll find some great advice in this episode, and as we know, networking is one of the key components to building out your career. And finally, we’ll talk about, Josh will give us some of his views on the grit and how important it is in your data science journey.
Kirill Eremenko: 00:02:53
Very exciting podcast. Get ready to be energized and pumped. Here we go. Without further ado, Josh Hortaleza, aspiring data scientist.
Kirill Eremenko: 00:03:10
Welcome back to the SuperDataScience Podcast, everybody. Super pumped to have you back here on the show. And today, we’ve got a special guest, Josh Hortaleza, calling in from Toronto. Josh, how’s the weather in Toronto today?
Josh Hortaleza: 00:03:23
It is pretty bleak. It’s overcast right now, so very happy to appear on this podcast to distract me from this inclimate weather. So, very happy about that.
Kirill Eremenko: 00:03:36
Do you get that crazy heat that’s going through the world right now since it’s summer in Toronto?
Josh Hortaleza: 00:03:41
Oh, yeah. Toronto is so variable when it comes to its weather. In the winters, there will be like negative 30 Celsius, and in the summers, it’ll be like positive 30 Celsius, so it’s quite drag to deal with it. But luckily, I work in technology, so I’m inside most of the day, and I’m very happy about that.
Kirill Eremenko: 00:03:59
Oh, good. Got you. Got you. How long have you been in Toronto for?
Josh Hortaleza: 00:04:04
My whole life. I was born and raised in Toronto, and I haven’t really left. I hope to do that soon, but with the pandemic, it’s not like those dreams can come to fruition any time soon, but that’s okay.
Kirill Eremenko: 00:04:20
And your background is from the Philippines and you’re planning on going back, checking it out. Is that right?
Josh Hortaleza: 00:04:26
Yeah, yeah. So, I was born and raised in Canada, but I feel that obligation to see where my parents have come from. The culture of the Philippines, even though I don’t culturally identify with it, is something that was a big formative experience for my parents having been born there. So, I think I owe it to my parents to visit where they grew up.
Kirill Eremenko: 00:04:47
Josh, I think from what I’ve heard and seen, Canada is quite a diverse inclusive culture, similar to Australia, where you have people from all sorts of backgrounds living, coexisting, and thriving together. Is that your experience?
Josh Hortaleza: 00:05:02
Yeah, absolutely. Toronto is a very multicultural city. It’s a very big economy as well, and it’s very nice to see this cultural mosaic of people who live in the city. And a big advantage of that is the food that you get to eat in Toronto. It is incredible.
Kirill Eremenko: 00:05:21
What’s your favorite food?
Josh Hortaleza: 00:05:23
My favorite non-Filipino, non-Canadian food is probably chicken tikka masala or something.
Kirill Eremenko: 00:05:29 Oh, wow.
Josh Hortaleza: 00:05:30
Yeah, yeah. I love Indian food. It’s really good. Granted, it’ll never be as good as wherever that comes from, like India probably, but it’s good enough for someone from the outside looking in. So, I’m very happy about that. But yeah, it’s a very diverse city. Very multicultural and it’s very reception to people who venture to that area. So I’m very happy to call Toronto my home in that sense.
Kirill Eremenko: 00:05:56
That’s awesome. I was speaking with my girlfriend, literally yesterday, about national foods, and she’s from the U.K., and we were debating which national foods come from where, and she was listing things from the U.K. like fish and chips, Shepherd’s Pie, roast, I think the roast meal, and then she’s like, “Curry.” I’m like, “Curry’s not from the U.K.” But it’s so popular here that you might as well think that it’s from here.
Josh Hortaleza: 00:06:28
Yeah. From the outside looking in, I was surprised to see that butter chicken was so popular in the U.K. I guess that’s a testament to your population, I guess your ever growing population. Right? So as the world gets more diverse, then our cultures get intermingled, and that’s very nice.
Kirill Eremenko: 00:06:48
Absolutely.
Josh Hortaleza: 00:06:48
That’s very cool.
Kirill Eremenko: 00:06:49
Absolutely. And what’s a national dish of Canada?
Josh Hortaleza: 00:06:52
It is poutine.
Kirill Eremenko: 00:06:55
Poutine?
Josh Hortaleza: 00:06:57
You know what? If you were to look at poutine and all it’s constituent parts and just look at them individually, you would say, “What is this? What’s this about? What is this?” In front of me, you’re like, “Gravy, cheese curds, maybe if you’re super Canadian, you’ll add maple syrup to it. I don’t.
Kirill Eremenko: 00:07:14
Oh, no.
Josh Hortaleza: 00:07:14
But you can add these things to it. But its parts are maple syrup. No, no, no. It’s gravy, cheese curds, and fries, and if you look at those, you’re like, “What is this?” But if you look at the other and you eat it, then it’s phenomenal. I love poutine, I am obsessed with poutine. It’s so good. I love all of it.
Kirill Eremenko: 00:07:35
It’s so delicious? Absolutely.
Josh Hortaleza: 00:07:37
It is phenomenal, it is amazing.
Kirill Eremenko: 00:07:40
Especially after a night out, after some drinks at a club, or music, or a bar, and you’re walking back and there’s all these foods that are bad for you, but full of fat and stuff like that, but if you have this dish, oh, it’s just so satisfying.
Josh Hortaleza: 00:08:04
I’ll satisfy like a day of cravings to feed my poutine addiction. It’s definitely something that I will risk taking to damage my diet in order to partake it. Yeah, it’s exactly how you describe. It’s a good comfort food, it’s good after a night out, and you can’t go wrong.
Kirill Eremenko: 00:08:23
Absolutely. Absolutely. Well, Josh, great to have you here. We met in the most random, in the literal sense, random way. Maybe I can describe the experience, but I’d love to hear it from you. If you can recap, how did that happen?
Josh Hortaleza: 00:08:41
So, it all starts with my buddy talking to me about, “Hey, there’s this online type of conference called DataScienceGO. And you’re into data science, so you should come along with me.” And I was like, “Yeah, sure,” and I did. And the conference was fantastic and we met through the booths. Not the booths, the working sessions that you had.
Kirill Eremenko: 00:09:03
This is the virtual right, DataScienceGO Virtual?
Josh Hortaleza: 00:09:04
That’s right. Sorry. DataScienceGO Virtual, that’s right. And there were these three minute break off sessions where whatever provider did the service that you hosted it on would match you with someone, and lo and behold, I was matched with you. And that was after I was religiously going on this thing and trying to meet as many people as possible.
Josh Hortaleza: 00:09:28
And one of the people who I met was you, and we talked for the three minutes of time that we were allowed to talk for and you asked more about my experience as opposed to me asking about your experience. So happy to come on the podcast to learn a bit about you.
Kirill Eremenko: 00:09:47
That’s awesome. Yeah. Man, it was a random thing. So for those who have worked at DataScienceGO Virtual, it’s like a lightening round of networking because virtual conferences usually lack that aspect. At DataScienceGO Virtual, we wanted to make sure we have that. And so, it’s a lightening round where you click this button, meet the next person, and for three minutes, you have a video chat with a random person from the audience, and then boom, time up. You meet the next person. That goes on for like 30 minutes several times throughout the event. So, tell us who else did you meet in these lightning rounds?
Josh Hortaleza: 00:10:21
I met some very cool people. I met people from the UCLA Anderson School of Business, I met someone who worked I think at NASA or something. That was pretty cool.
Kirill Eremenko: 00:10:32
Wow.
Josh Hortaleza: 00:10:32
Yeah.
Kirill Eremenko: 00:10:33
Oh, I think I met her as well.
Josh Hortaleza: 00:10:34
Yeah, yeah. It was not that she’s not there anymore, but she was working somewhere else then, but that was pretty cool. I met someone who worked for National Grid and I met multiple people. And I actually met with them after the actual conference. I actually had hours long conversations with all the people who agreed to meet with me after.
Kirill Eremenko: 00:11:01
Go, Josh. That is epic. Speaking of dedication and follow through, that’s awesome man.
Josh Hortaleza: 00:11:07
Yeah. And because the conference was international, I got to get a lot of perspectives from the American space and perspectives from places really outside of my scope, which is the Canadian scope or the Ontario scope. So, very fortunate that people were able to lend me their ear and talk to them about their passions I guess, right? So, I was very happy and benefited very much from doing the lightning rounds.
Kirill Eremenko: 00:11:34
Fantastic, man. Fantastic. Hope you’re enjoying this amazing episode. We’re going to break for a quick announcement, and it’ll be straight back to it. This episode is brought to you by our very own virtual data science conference called DataScienceGO. If you haven’t been to DataScienceGO yet, if you haven’t heard of DataScienceGO, check it out at DataScienceGO.com/virtual.
Kirill Eremenko: 00:11:56
There you’ll see a recap of the incredible event we had in June this year. We’re hosting DataScienceGO Virtual number two in October. Make sure to be there. We’re going to have amazing speakers, amazing energy, and we’re going to have virtual networking three minute sessions to connect with your peers, with mentors, with hiring managers, with mentees, with whoever is at the conference random lightning networking three minutes each. You can stay in touch with these people, expand your data science network, be there. It’s in October. It’s absolutely free. The best part, it’s absolutely free. Just go to DataScienceGO.com/virtual, register for the event today. All right, let’s get back to the podcast.
Kirill Eremenko: 00:12:40
Well, absolutely cool. So, we met and now we’re here and the reason why I invited you is, when we were talking just for those three minutes, I could feel so much, I don’t know, that you’re super smart and you are determined to go where you’re going and you’re very curious. Right? So, putting all those three things together, I thought I’ve got to talk to you, and yeah, we’ve quite a few things to cover. So before we get started, give us a bit of background. Who are you? And what is your life stage right now?
Josh Hortaleza: 00:13:15
Awesome. So my name if Josh Hortaleza. I am 22 years old. I’m from Toronto, Ontario. I am an enrolled third year student studying computer science at a school called Wilfrid Laurier University, go Golden Hawks. I kind of stumbled. As to data science, I kind of stumbled across data science. I describe it as I tripped over it one day and I discovered it, and before that, I actually wanted to get into equity research.
Josh Hortaleza: 00:13:45
That’s like Hedge Funds and investments and like that. And I was actually quite good at delivering investing and pitching stocks. I actually won contests in that area, and I intended to go into the equity research space. And the thing is, and even though I studied computer science, I wanted to do this business thing so bad. But one day I went to a workshop posted at the nearby university close to me, and I was kind of introduced to this field of data science.
Josh Hortaleza: 00:14:20
And I was like, “Wow. Okay, this is really, really cool. I think this is very fascinating.” And before ever venturing into this space, I only heard of data science from like, “Oh, this is an area that when we pitch, he’s going to. This is where the Google algorithms come into play and what not.” So it’s like, “Okay. Maybe this thing, maybe I couldn’t do this.” So I went through this workshop, I did the project, and then suddenly I got a job as a data scientist intern at a company called CIBC, and that’s a bank in Canada.
Kirill Eremenko: 00:14:53
What do you mean suddenly? How do you suddenly get a job? Did you like wake up and you have a job?
Josh Hortaleza: 00:14:57
Yeah. It was like, “Wow. This secured us.” But no, I was definitely having a hard time looking for a job back when I didn’t have a lot of experience. And when you’re a student, you kind of have to get the ball rolling in some ways. And this is especially hard when you are trying to pursue two areas at once. So, I discovered the data science as a whole in around February of 2019. And then, I got my first job in data science on May 2019. So, it was very-
Kirill Eremenko: 00:15:37
That’s fast.
Josh Hortaleza: 00:15:38
Yeah, it was very fast. Well, first of all, I widened my net. I was willing to work anywhere, right? So I would work in Toronto, I would work in the areas around Toronto, I would work across the province. And I ended up working in a place called Waterloo, which is where my school is, so that was fine. And I decided to broaden out my options if possible, so I applied to data science jobs in any context.
Josh Hortaleza: 00:16:03
So that’s in insurance, that’s in capital markets, that’s in eCommerce, that’s in healthcare data science. So, I really expanded my options and by pure virtue of numbers, I just was bound to get one, I guess. I’m not bound to get one because it doesn’t come to that, but I consider myself very fortunate and lucky to get the job that I have, but ever since then, I just kept going. I found this job through applying on a niche job board and that’s how you get the ball rolling in that space.
Kirill Eremenko: 00:16:37
What’s a niche job board?
Josh Hortaleza: 00:16:40
So it’s exclusive to the Waterloo area. It’s called Work In Tech, and that’s run by one of the incubators in the area, so they’re the startup incubator. It’s unfortunate that not everyone can go on it, because it’s a Waterloo exclusive. But yeah, that’s where I found it. Very niche.
Kirill Eremenko: 00:17:01
Okay. But people will have local ones in their cities, or something similar?
Josh Hortaleza: 00:17:05
Yeah.
Kirill Eremenko: 00:17:07
Got you. Okay. So, you got this job. I can see from LinkedIn, you stayed there for four months. What happened next?
Josh Hortaleza: 00:17:13
In terms of what happened at the job? Or where I worked next?
Kirill Eremenko: 00:17:17
Whatever. Whatever is worth talking about.
Josh Hortaleza: 00:17:21
Yeah. So I think both of them are worth talking about. At CIBC I worked a lot on NLP tasks. So, the retail banking space has a lot of data and has a lot of text data. Why? Because a lot of the value add or the business driver of banking is in the retail banking space. And particularly what I worked on was call center feedback. So what essentially that is, is okay, well there are call centers that people who sell cars or provide customer support and then they are given a survey.
Josh Hortaleza: 00:17:56
And they’re asked for their feedback. And we take that survey data and feedback data and then we perform NLP stuff on that. So in my case, it was sentiment analysis. So, my end deliverable was kind of a dashboard that went through these thousands of feedbacks and put them in buckets from a bad experience, neutral experience, and good experience. And we used Google Dashboard for that or something. We used a dashboard or service for that, and yeah, Google cloud platform. Yeah, GCP. So, once we had this deliverable, then we presented our findings to management and we tried to help them intelligently improve their feedback from that.
Kirill Eremenko: 00:18:39
Okay. Tell me, two questions. One is, what techniques did you use for the NLP?
Josh Hortaleza: 00:18:46
So to be quite honest with you, we used a lot of APIs. So, we used spaCy and we used SKlearn for the second one. Yeah, so we used these things and then we just invoked all the sentiment command and we just applied it to all of the-
Kirill Eremenko: 00:19:06
So plug and play, right?
Josh Hortaleza: 00:19:07
Yeah.
Kirill Eremenko: 00:19:07
The evidence, it’s not a surprise. And like you started into data science in February. Nobody’s expecting you to be a NLP wizard by May. It’s three months. So mostly products, right? You were using to get the insights?
Josh Hortaleza: 00:19:26
Exactly. But the thing is, people who work with me had graduate degrees and what not. So, we were all working on the same thing. I think it was a matter of what the bank needs at the time. And a bank doesn’t necessarily need the latest and greatest model that’s coming out of academia, right? They can just use some rudimentary thing to provide some value in that way.
Josh Hortaleza: 00:19:52
Even though I was very happy with the experience of working there, but it wasn’t rigorous to me, I guess, for lack of a better word. But the next internship was at Intapp Insurance and that’s in Toronto. And the job was also called data science. I was a data science intern there. And the reason why I want-
Kirill Eremenko: 00:20:15
Sorry. Before we jump to that one, the second question I wanted to ask you, at CIBC, what, if you can share of course, I don’t want to overstep into any kind of confidentiality parts, but what kind of insights were you driving with these dashboards to the executives?
Josh Hortaleza: 00:20:33
Let me think about it. Yeah. So essentially, the insight they were trying to drive, “Okay, this is what your customers are experiencing.” And there’s a business team that would take this data and try to interpret stuff from it, interpret insight from that. And then, we would work with them to do that because we are the translators of data. Right? We try to bridge the gap between the technical and the business. Right? So, we try to help them do this. And I guess I can’t really delve in on the actual-
Kirill Eremenko: 00:21:03
No, no, no. I understand, I understand. Yeah.
Josh Hortaleza: 00:21:05
Yeah. But it was a very cross functional process. Right? So we worked with a lot of stakeholders, we worked with a lot of people like that, and then we presented this thing. But yeah, that’s kind of how that was.
Kirill Eremenko: 00:21:17
Okay. No worries. Let’s move on. Second internship?
Josh Hortaleza: 00:21:19
Yes. So second internship was at a place called Intapp Insurance and this was surprisingly, but probably … This was a fantastic internship. I love this job. I love the people who I work with.
Kirill Eremenko: 00:21:34
I can feel it in your voice how excited you are.
Josh Hortaleza: 00:21:36
I gush about Intapp all the time. And the reason why I value it as a fantastic internship is because, number one, I got to work with incredibly very cutting edge models. So I can give a bit context on that. I worked with CNNs in an image processing use case.
Kirill Eremenko: 00:21:59
Convolutional neural networks, right?
Josh Hortaleza: 00:22:01
Yes, that’s right. Convolutional neural networks, yeah. So I did a lot of image processing stuff. So to elaborate, in the insurance space, there are a ton of forms that you needed to fill when you submit a claim. So the thing is, how it’s done in the past is it’s done manually. You would classify all these documents manually and then you would send each paper or form to its appropriate division that handles that document. Right?
Josh Hortaleza: 00:22:32
This can be automated and this can be automated with a model. Right? And the model, being a CNN model, and then if you use the model to do what the manual classification takes and you can do it much faster. So, it was very cool internship because I used CNNs to do this image processing. And I got to read some very interesting papers. Like a-
Kirill Eremenko: 00:22:57
Sorry, sorry. Let’s recap, just so I understand better. So, somebody fills in this form. It’s a form they fill in by hand, and then they send it in and you take a scan of the form and then you need to get information out of it. Is that right?
Josh Hortaleza: 00:23:13
So it’s more like you have to identify what the form actually is and then put it in the appropriate bucket. So I guess the forms that exist are like legal forms, police forms, segregation forms, photos of the damaged site, stuff like that, and-
Kirill Eremenko: 00:23:28
Okay. And they’re all scanned in and you just need to find the title of the form using CNNs?
Josh Hortaleza: 00:23:36
Not necessarily. The title isn’t necessarily there.
Kirill Eremenko: 00:23:39
Ah, okay.
Josh Hortaleza: 00:23:40
So it’s interesting in that respect, because the thing is, you can have a form that looks like a standard form you could fill out, and then you can have a picture. Right?
Kirill Eremenko: 00:23:48
Uh-huh (affirmative).
Josh Hortaleza: 00:23:48
But the picture and a form could be classified in the same bucket. Right?
Kirill Eremenko: 00:23:53
Uh-huh (affirmative).
Josh Hortaleza: 00:23:54
Because they pertain to the same thing, right?
Kirill Eremenko: 00:23:55
Uh-huh (affirmative).
Josh Hortaleza: 00:23:56
So it was very interesting to delve into the actual raw data to see, okay, where is the CNN failing? What can I do to make the process better? How can I improve the model in some way? So yeah, that’s what I did there.
Kirill Eremenko: 00:24:12
Got you. Okay. All right, good. Yeah. And so, what did you enjoy the most about this one?
Josh Hortaleza: 00:24:16
So I mentioned two things. The first part was that I got to work with very important, or not important, but very cool technology. And in doing that, I actually went to a lot of conferences. So in Toronto, Toronto is the home of Geoffrey Hinton and-
Kirill Eremenko: 00:24:36
Yeah, the Godfather of AI.
Josh Hortaleza: 00:24:38
Yes, the Godfather of AI. And Intapp paid for me to go to a conference that was commemorating his Turing award. So I got to-
Kirill Eremenko: 00:24:49
Amazing. You got to see him speak?
Josh Hortaleza: 00:24:51
I got to see him speak and I got to see-
Kirill Eremenko: 00:24:52
No way.
Josh Hortaleza: 00:24:53
Yeah.
Kirill Eremenko: 00:24:54
Man, that’s awesome.
Josh Hortaleza: 00:24:54
It was really cool. And people who were there were, like Ilya Sutskever who is the OpenAI guy, the guy who does OpenAI. So he-
Kirill Eremenko: 00:25:02
Okay. I haven’t heard of him, but also I know about OpenAI.
Josh Hortaleza: 00:25:05
Yeah, yeah. So he’s the chief scientist at OpenAI, and he was the guy who beat the Dota teams with his neural network over there. And that was amazing to meet that guy. But Geoffrey Hinton, listening to him talk and go through the history of a neural network was fantastic. And it was really interesting to see what a venerated figure looks like in this field. Because this field is very young, right? This a very young space. But he’s been working on it for a long time.
Kirill Eremenko: 00:25:39
Since the 1980s, right?
Josh Hortaleza: 00:25:41
Yeah, absolutely. Even before that, it was the Saffron or something.
Kirill Eremenko: 00:25:47
Okay, yeah. Yeah, yeah.
Josh Hortaleza: 00:25:47
And Bapta, back at that time. So to watch him talk about the adoption of AI, its adoption of neural networks right now even though he’s been doing all this work, him and Yoshua Benjio and all those.
Kirill Eremenko: 00:26:00
Yann LeCun.
Josh Hortaleza: 00:26:01
Yann LeCun, these pioneers. Aaron Courville. All these people who work on this thing. Even though they’ve been working for a long time, it’s only now reaching global adoption. And that was I think after … I think the revolutionary thing was called AlexNet. And it was in the image contest where they’re like, “Wow, this thing beats all of these other things. So, we’ll adopt this thing.”
Josh Hortaleza: 00:26:26
And insight of an important story he brought up was this notion of a continental drift or something. And people at the time of the theory being brought about, they were like A, this isn’t true, we’re going to mock you for this. And this conjecture, are you crazy? How can continents move? But now we know that continental drift, the moving of these large land masses is an actual thing.
Josh Hortaleza: 00:26:59
But at the time of Dr. Hinton’s early career, the neural network didn’t really get adopted, right? People were laughing this thing out of the conference rooms and laughing this thing out of these journals. But now, it’s like widespread. It’s accepted that this works. So, listening to him talk about the history of the neural network was fantastic. So, that’s that.
Josh Hortaleza: 00:27:30
And I guess the second thing that I learned from Intapp was how to be a good worker. And how to be a value add to a company. They had three one on ones with me every single week. I learned a lot about myself as a worker and as an employee. Especially where I fit in a company and what value add data can bring. So, I was really adamant in trying to figure out how do I be a good data scientist and what does that mean? What does it mean to be a good data scientist in a company context? And I learned that there.
Kirill Eremenko: 00:28:05
Amazing, go Intapp. Love it.
Josh Hortaleza: 00:28:07
Yeah, that was fantastic. I really cannot sing the praises of Intapp enough. I love working there. But the thing about Intapp was it was a four month thing. So low and behold, I had to leave. But it is what it is, right? And where I’m working now is a startup called Cognitive Assistance Corporation. And what they do is really cool. They do wifi motion technology. They use wifi to detect motion.
Kirill Eremenko: 00:28:40
Wow.
Josh Hortaleza: 00:28:41
And I can’t really get into the work I’m doing there because it’s really research-
Kirill Eremenko: 00:28:47
Yeah, yeah. Trade secret.
Josh Hortaleza: 00:28:48
Yeah, it is. But seeing how a startup grows from the ground up, you realize it’s a whole different ballgame working at a startup as opposed to working at a big company. And this company that I’m working at is really on their game. And I see their office and there’s tons of computers lying around, tons of work, physical wrenches and pliers and stuff. Because they’re really trying their best to work on whatever works, right?
Josh Hortaleza: 00:29:22
At these big companies, they have a really formalized process. But at these startups, they’re like, “Okay, if it works, it works. And then, we’ll go with that.” Not to say that that’s an unintelligent process. That’s a very smart way to go about utilizing intelligent people. So, I’m having a great time working there right now. That’s kind of my career background.
Kirill Eremenko: 00:29:45
Man, incredible. At 22 years old, you’ve had the experience of big corporations like banking and you’ve had experience at startups. Dude, you’re going to go places for sure. Mark my word.
Josh Hortaleza: 00:29:57
I actually worked two internships before that, before CDIC. But I’m not going to discuss that because that’s in the realm of finance. But yeah, I’ve worked almost five internships right now. I love working, as you can see. I love working and I love learning. And I’ll get into why I do that later if you want.
Kirill Eremenko: 00:30:19
Gotcha, gotcha. Your internships interestingly have all been four months, four months, four months, four months. And now at Cognitive Systems Corps, you’re on month number three. So, are we going to see the same pattern? Are you going to be four month or you think you’re staying longer?
Josh Hortaleza: 00:30:35
I’m going to stay for four months. I generally have a rule where I will only be at a company for four months as an internship. And the reason why that is is because I figured the vast majority of the things that you could learn happens in the first four months of a company, right?
Kirill Eremenko: 00:30:50
Wow.
Josh Hortaleza: 00:30:51
So, learning about the company culture, learning about what I want to do, learning about what they use and what not. I feel like to broaden your horizons is the best approach you can undertake when you’re in the learning process, right? Because not only are you learning about what companies use data science for, but you’re also learning about what you believe is good company culture and where you fit into this whole economic system. So, that’s why I jump from company to company because I’m trying to find my niche. I’m trying to find what I really want to do I guess.
Kirill Eremenko: 00:31:30
Interesting. That is incredible. That is one of the most audacious and at the same time valuable pieces of advice I’ve heard on this podcast in terms of building your career. Starting your career. So, it’s not just like a coincidental thing, it’s by design that you’re doing these internships and you’re not actually seeking a full-time job right now. You’re looking for all this experience. Is that right?
Josh Hortaleza: 00:31:56
Yeah. I’m still in school. And I’ve kind of delayed school to undertake this. And I think it stems from some form of imposter syndrome where I feel like how do I know what I know enough, right? Because I only entered this field in February of 2019. That isn’t a long time, right? I definitely don’t know everything and everything there is to know about the field. So the question that I’m trying to understand is when do I know that I’m good enough for the full-time market, right?
Kirill Eremenko: 00:32:29
Okay. Interesting.
Josh Hortaleza: 00:32:32
That’s another component of me jumping around, I guess.
Kirill Eremenko: 00:32:35
Got you. Have you heard of the 37% rule?
Josh Hortaleza: 00:32:38
I have not. What is that?
Kirill Eremenko: 00:32:39
It’s a mathematical solution to how many people you need to date or how many people you need to not settle down with before you actually start considering whom you will get married with. And basically, I will include in the show notes, if anybody wants to check it out, it states that in order to get a statistically significant sample and make a right choice about who you’re going to settle down with, you need to take the total number of people you think you can potentially date in your life.
Kirill Eremenko: 00:33:15
And first of all, let’s say it’s 100, hypothetically, right? Then the first 37% of them, in this case 37 people, you definitely cannot settle down with because then you’re going to have a too high a chance of missing out on the right person. You don’t have enough. Pretty much you’re doing the same thing. You need to calculate what’s the maximum number of jobs I can have in my life and take 37% of that and discard the first 37.
Josh Hortaleza: 00:33:41
Exactly. And it’s tough to do when the horizons are so broad. I wanted to do equity analysis and now I’m going to do this. And now, I’m very interested in consulting, right? So, maybe at Deloitte or something or what not. That’s all very interesting to me. And it feels like there’s only so much you can do in so little time. So, I have to pick my next opportunities very intelligently to try to pass as wide of a net as possible in my intellectual journey, right? And so, I have to see what companies fulfill multiple buckets that I want to go through next, right? So, yeah that’s my process.
Kirill Eremenko: 00:34:23
Here’s a question for you. So, not only the horizons are broad, but also the field is changing all the time. And you said how do I know that … I don’t know everything about the field right now, I need more internships, more experiences. But the problem I see with this is you will … Yes, by more internships you will get to know the field more, but let’s say your next internship is going to be reinforcement learning and then RPA and then I don’t know, more NLP.
Kirill Eremenko: 00:34:51
By the time you do those internships, the field of CNN would’ve moved forward. And now, you have to do another one in CNN. So like, where is the end with this? Because the field is growing all the time.
Josh Hortaleza: 00:35:00
That is a fantastic question. What I have to reconcile eventually is that I am good enough. And that’s the thing. Because in the back of my head, I know that you hire a data scientist not because they know something but because they have the potential to know something. Because the field changes so fast, right? This is a constantly changing field. All of my jobs that have been called data science I have done radically different things at.
Josh Hortaleza: 00:35:28
So, the actual field is changing. Therefore the field wants people who can learn to adapt to that change and thrive off it. So, I guess even though I’m on this journey right now, the journey ends when I come to the conclusion that I am in fact good enough. And when that comes along, I don’t know. But in order for me to graduate, I have to understand that within myself. So, I do a bit of soul searching I guess. And once I truly believe in myself, then I’ll enter the full-time market and then we’ll see. That’s when it stops I guess.
Kirill Eremenko: 00:36:06
What do you say to a psychologist that would tell you right now, right here that you are good enough. That everybody’s good enough just because we’re human. Regardless of our backgrounds and experiences and securities, we are all good enough. Why do you not stop right now?
Josh Hortaleza: 00:36:24
Do you mean intrinsically by virtue of being a human being?
Kirill Eremenko: 00:36:27
Yeah.
Josh Hortaleza: 00:36:28
I mean, yeah, I see that. But we live in a system where you have the beauty of choice I guess. So, I can believe I’m good enough. I can definitely say that, “Hey, by virtue of being a human being, I am good enough to …” If I’m putting this correctly, “I am good enough even with a high school diploma,” right? Or something.
Kirill Eremenko: 00:36:54
If you have one.
Josh Hortaleza: 00:36:56
Yeah, exactly. And that’s true. You can probably do this job, do some jobs in data science without a lot of significant education at a university. However, even though that could be the case, what I would say is that I do not know how other companies think. Because I am not a corporation, I’m a person. So, I cannot leave it up … If the other person who’s sitting across the interview table from me is also thinks that, then boom, that’s awesome.
Josh Hortaleza: 00:37:33
Maybe I have a chance, right? But this is such a media filled field. Everyone is obsessed with machine learning, people think it’s like amazing and what not. And it is. But the mediazation of machine learning and artificial intelligence, especially as it’s so visible with companies like Facebook and Amazon and what not, it’s the case that I cannot leave my employment and my future career to chance. To say that, “Hey, I can’t assume that the person sitting across from me thinks that as well.”
Josh Hortaleza: 00:38:07
Even though I generally do believe that. It’s all about intrinsic knowledge and what not, and by virtue being there you go, you could probably do it. What I would say to the psychologist is that I cannot leave my career up to chance. And I try to limit the luck factor as much as possible.
Kirill Eremenko: 00:38:27
Got you, got you.
Josh Hortaleza: 00:38:27
Because I do not know how my interviewer thinks. And therefore, I have to make sure that … Kind of persuade them that I’m already a finished product.
Kirill Eremenko: 00:38:44
Yeah, got you, got you.
Josh Hortaleza: 00:38:44
So, that’s what I would say to that. You can be right, but something that … It doesn’t mean that it’s right in every context, right? So, that’s what I think about that.
Kirill Eremenko: 00:38:59
Got you. Interesting. Some might say that rather than limiting the luck factor, check this out. There’s a definition of luck as when opportunity meets preparation, right? Opportunities come along, but if you’re not prepared, no luck. So, some might actually say you’re maximizing the luck factor by preparing and doing all these internships, going through all these different parts of data science, different fields in data science to when the right opportunity comes along, you are prepared.
Josh Hortaleza: 00:39:29
Okay. So, I’ll rephrase what I said before. But when I mean luck, I mean random chance, right?
Kirill Eremenko: 00:39:33
Random chance, got you.
Josh Hortaleza: 00:39:34
So, I totally agree with you that luck is by design, right? I had to put myself in a position to be ‘lucky’, right? The thing is my job right now at Cognitive, I got through an opportunity that would be considered lucky. Because the thing is, I actually lost my other internship that I was supposed to do because of the Corona pandemic. I was supposed to be an equity analyst at a hedge fund, right?
Josh Hortaleza: 00:40:07
But that fell through because of Corona and I had to scramble to get a job. In my networking, one of the people who I knew for over a year and a half, I reach out to him and this is how I got the job I have now, right? But the thing is, I wouldn’t have had that opportunity if I didn’t know what I was talking about from working at Intapp and CIBC. And I would have never have met him if I didn’t actively network beforehand, right?
Josh Hortaleza: 00:40:38
So, I definitely am a big believer in putting yourself in a position to be lucky. But I guess before I met … I don’t want to chalk up my career to random chance, right? So, that’s all I’ll say about that.
Kirill Eremenko: 00:40:54
Okay, got you, all right. Right now, before the podcast you told me your dream is to break into the field of fundamental machine learning research and corporations. The likes of Facebook AI, Google Brain, DeepMind and so on. Tell us, is that a dream or a goal? Difference being a dream is something you don’t have a timeline on and a goal is something that you’d have a timeline on.
Josh Hortaleza: 00:41:22
It’s a goal, it’s a goal. I have dreams but they have nothing to do with my career. I don’t dream about a career. But it is a goal. I would consider it a goal, yeah. I would definitely describe it as a goal, yeah.
Kirill Eremenko: 00:41:40
What steps are you taking towards this goal?
Josh Hortaleza: 00:41:41
Understanding what it takes to get there. Understanding the different paths to get there, right? And I do that through networking and I do that through reading what people have online. I read a lot of blogs. And I try to upskill myself to try to match that profile, right? I obtain a fundamental understanding of what the job generally looks for, I try to match up to that criteria. And then, I try to once again maximize my luck by expanding the opportunities I can get through networking and applying to jobs. So, that’s kind of the steps I’m taking right now to enter that space.
Kirill Eremenko: 00:42:26
Got you. We’re going to talk about networking just now. I want to make a plug for you man because I believe that talented, driven people who have a vision of where they’re going deserve to get to where they’re going. So, if anybody listening to this is in the space of fundamental ML research in Facebook AI, Google Brain, Uber, DeepMind, a Twitter Cortex, whatever it is, or you know somebody there or you work there, you know a manager there, send them this podcast. Send them this podcast, tell them about Josh and let’s help Josh get to his goal faster.
Josh Hortaleza: 00:43:04
I appreciate that, thank you.
Kirill Eremenko: 00:43:07
Man, my pleasure, my pleasure. But let’s talk about networking. You mentioned networking a couple times, why is this such an important component of your career development?
Josh Hortaleza: 00:43:17
Awesome. I’ll talk about my own personal experience and then I’ll talk about why it’s important in the grand scheme of things.
Kirill Eremenko: 00:43:27
Sounds good.
Josh Hortaleza: 00:43:28
My personal experience in networking is I used to want to do finance and in business school, networking is huge, right? You learn the tricks and you learn the tips and you get a lot of experience just talking to people, right? And you do all of these thing. And I networked a ton to get into the industry, to learn about the space. But in my time in networking, my error when I was entering finance is that I kind of had a short-term approach.
Josh Hortaleza: 00:44:05
Where I kind of expected a job right away. And that comes from an ignorance of not understanding what networking is. Networking, now that I’m a bit older, I understand that networking is a full experience that goes across many years, right? Or building these relationships and what not. And building relationships that are strong and that are real and that are not inmaterial. And what I mean by inmaterial, it’s not like I’m talking to you only because I want a job, right?
Josh Hortaleza: 00:44:42
I want to talk to you because I want to learn from you and maybe you can learn from me and maybe we can share information. And maybe we can help each other. And eventually, could be, maybe if the opportunity ever arises, then I can get a job that way, right? That’s what I’m adopting right now. And now that I’m kind of outside of the finance space, I kind of get a bad taste in my mouth whenever I hear networking because people who don’t understand it will be like, “Okay, I want to use this to get a job right away,” right?
Josh Hortaleza: 00:45:14
And when you do that, you don’t see a person as a person, you see them as a means to an end, right? It’s a very inhuman thing to do. It totally lacks any sort of compassion and human emotion.
Kirill Eremenko: 00:45:28
And people pick up on that, right? People have perfect BS meters, right? Or sensors.
Josh Hortaleza: 00:45:34
They do, they really do. They can smell it a mile away. Especially when you’re a kid coming out of university who doesn’t know how to talk at all, right? Like this is a whole different university, from high school it’s a whole different game, right? Now you’re no longer dealing with your peer group, you’re venturing outside of your peer group to people how are more savvy than you are, right?
Josh Hortaleza: 00:45:54
You think you’re so smart in high school but when you enter university and expand your horizons, you realize that you’re not as smart as you think you are. No, you’re not. You’ve just left your bubble, right?
Kirill Eremenko: 00:46:06
So, what’s your solution? I don’t want to say trick or tip, but how do you network in a way that’s genuine?
Josh Hortaleza: 00:46:17
You don’t think about networking at all. I mean, of course in the back of your mind, but when you’re talking to someone, you talk to someone because you are genuinely interested in what they do and the experiences that they have, right? You’re not talking to them because you want a job. You’re talking to them because they’re willing to share their human experiences and perspectives with you and you are interested in that, right?
Josh Hortaleza: 00:46:46
So, if you were to take a blanket approach in networking, like firing off LinkedIns all the time, then chances are you wouldn’t really be interested in every single person you’re talking to. Only the company they’re at. And they can tell if you’re not interested in what they’re doing. So, when I try to talk to someone, I talk to people who I’m genuinely interested in talking to.
Josh Hortaleza: 00:47:11
Another thing with networking is that you want to be a good listener. So, the reason why you network is because you want to listen to the perspectives of someone else. However, in your feeble attempt to get a job right there and then, you will talk about all of your cool stuff and all of your cool experiences. And they may be good, they may be very strong experiences, however you are not there to talk about yourself. You’re there to listen.
Kirill Eremenko: 00:47:40
Or you will listen but you won’t hear them because all your reticular activating system in your brain is searching for opportunities to like, “Oh, can that serve me? Can that help me get a job or whatever else?” So, you’re focusing on the wrong things in the words that they’re saying. And that’s not going to lead to anything.
Josh Hortaleza: 00:47:58
Exactly. When they’re talking to you, you’re thinking about the next thing to say as opposed to thinking about what they’re saying. However, that’s counterintuitive because if you listen to what they say, you become an active listener, right?
Kirill Eremenko: 00:48:10
Absolutely, yeah.
Josh Hortaleza: 00:48:12
And you get to have a more engaging conversation by bouncing off of what they’ve said to you and then you become more genuine way. What I try to do now as a networker, as someone who is networking is dismantle the notion that I am a parasite, right? Because immediately when you’re networking, you have your guard. If some kid asks to talk to you, you’re probably going to have your guard up. Be like, “What does this kid want?” Right?
Kirill Eremenko: 00:48:42
Yeah.
Josh Hortaleza: 00:48:43
Or you know, “Oh, I know exactly what this kid wants. And I’m just going to play with him for a bit just to see what he’s about,” right?
Kirill Eremenko: 00:48:50
Tell me, how do you do that? How do you dismantle that notion that you’re a parasite?
Josh Hortaleza: 00:48:55
Yeah. How I would do that is by being a genuine person because technically speaking, not to be cynical or anything, but you can still absolutely be a parasite and be genuine, right? You can be a genuine parasite, right? So, there’s that. But I try to be as genuine as possible because I’m genuinely interested in their story.
Kirill Eremenko: 00:49:22
But tell me, let’s say you met some people at DataScienceGO Virtual, you only spoke to them for three minutes, you exchanged contact details. What’s the next step? Indeed, everybody is interesting. That’s the foundation that is important to understand. Everybody inside is interesting, they have something cool to share, right? But how do you approach them afterwards to actually learn what they have to share? What do you say to them?
Josh Hortaleza: 00:49:48
I say to them, “Hey, I really liked our conversation at DataScienceGO Virtual, even though it was three minutes long and that’s regrettable. I want to continue that conversation more because I’m very interested in what you’re about.” Right? That’s how. Continuing the conversation, that’s what you want to do. You want to continue the conversation. So people who I talked to from DataScienceGO were willing to continue that conversation with me and we just talk about things that interest us.
Josh Hortaleza: 00:50:19
And that’s how these genuine interactions happen. Whenever I talk about this, I kind of think about how did you meet your friends, right? When you think about that, or how do you make friends, right? And when asked this, people think about, “Hey, how did I meet my friends? Did I deliberately try to go out of my way? Did I have some sort of robotic approach?” Step one, talk to them. Step two, do this. Right? No, no, no.
Josh Hortaleza: 00:50:48
They just happen to be your friend. They just happen to met them in some artificial environment space that’s school or work or university or not. And then, by virtue of talking to each other and sharing similar interests, then there you go, you’re friends now. Right? And that’s how it is in networking. But you just go in with the maybe approach that you could get something out of this. But you have to go in like a genuine person, right?
Josh Hortaleza: 00:51:15
So, on a applied basis, or to say something that’s more applied, I would say okay once you met them for the first time with some medium in which to do that. One of them would be DataScienceGO Virtual. Then you would ask to continue the conversation later. And when you’re doing that, and when you’re actually doing that, you want to try to yes, be genuine. Yes, if you’re genuinely not amazing at social mores, your goal is to be likable, right?
Josh Hortaleza: 00:51:51
Your goal is not to impress. Your goal is to be likable. Your goal is to be a human being. Your goal is to be someone who they can sit on … They can wait on an airplane. Let’s say you’re in an airplane terminal and the airplane is coming in five hours, right? Your goal is to be able for them to say, “Hey, I can talk to this guy for five hours at this airplane terminal waiting for an airplane,” right? That’s what you want to try to do.
Kirill Eremenko: 00:52:19
Got you. Have you read How to Win Friends and Influence People by Dale Carnegie?
Josh Hortaleza: 00:52:23
I have the book, I have not read it yet. It’s regrettable.
Kirill Eremenko: 00:52:26
Man, you’ve got to read it. I recommend to everybody. The things you’re saying right now, it’s a matter of experience and knowing how. For some people it comes naturally. For me for instance, I’m not naturally. I’m more robotic naturally, I’m more introverted and so on. And so, this book really outlines how to be an active listener, how to listen to people, how to respect people, how to be curious in a genuine way about what people have to say.
Kirill Eremenko: 00:52:57
Fantastic book, I recommend for everybody to check it out. It ties in very well with the things you’re talking about. Because then even reading the first three chapters, you’re already going to get so much value. It will be like night and day in terms of your ability to do exactly what you’re saying.
Josh Hortaleza: 00:53:14
And so, that’s my experience. And I guess the second reason why I would consider that important, and I agree, the book is a fantastic book. I have to read it. I’ll read it soon, for sure. But I guess the second reason why networking is important is because you never know what’s going to happen, right? You never know, right? To have this core pandemic happen, boom, there you go. It came out of nowhere, right?
Josh Hortaleza: 00:53:38
If you’re an average person, Corona virus came out of nowhere. And suddenly, you’re stuck in your house for 18 hours a day, right? So, networking would help if you lose your job or something. Like now, you have a way to occupy your time or different channels in which you could get a job again, right? Because you never know, right?
Josh Hortaleza: 00:54:00
I had someone describe networking to me as planting seeds and growing a relationship I guess. And eventually picking the fruit later, right? It’s important for your professional career, number one because you never know what’s going to happen in your career. Like something totally bad could arise that results in you losing your job. And someone could help you later. And number two because it’s important to know what other people in your space are doing and are learning right now.
Josh Hortaleza: 00:54:35
Because you have to … Especially in the field of data science, you have to be constantly learning. So to get an understanding of what your peers are doing in this space, is fantastic. It’s phenomenal. So, that’s kind of why networking is important.
Kirill Eremenko: 00:54:48
Absolutely, absolutely. And you never know. Maybe you will help someone, you know? You’re networking with someone and so on and then they lose their job and you know somebody at your university or somewhere else. And you can connect people. There’s also this element of giving back. If you go into it with, “Hey, even though I might not have all the experience in the world, I might not be able to get people. I don’t have a company to hire people into right now. I still might be able to help people this way.”
Kirill Eremenko: 00:55:16
Once you have this attitude like what do you bring to the table, how can you help people? And that potential for it is there. I think it also adds to the element that you’re not just doing it only for selfish reasons.
Josh Hortaleza: 00:55:31
Exactly, yes, exactly, yeah.
Kirill Eremenko: 00:55:33
Mm-hmm (affirmative), mm-hmm (affirmative). Got you. Okay, all right. Josh, it’s been a blast. I can’t believe it’s already been 50 minutes that we’ve been chatting. There is so many more things I wanted to talk about but it’s been amazing. I think we’ve got to start wrapping up. Definitely we should have a follow up episode. Maybe sometime in a year or two when you’ve done another 20 internships.
Josh Hortaleza: 00:55:57
Absolutely, yeah.
Kirill Eremenko: 00:55:58
Before we wrap up, I want to ask you among all the things we’ve talked about, maybe there’s something burning, a burning piece of advice that you want to share with people that we just didn’t have time to cover. There’s plenty of other stuff I wanted to ask you about, but nevertheless, what’s the one extra thing you’d like to share with people out there who are in a similar position to you, breaking into this field of data science? Who have a dream, who have a vision and are excited about the field, what would you say to them?
Josh Hortaleza: 00:56:28
You have to be gritty. You have to be gritty. The job market is a tough beast. It’s very frustrating. And from a student’s perspective, I can only speak from the perspective of myself as a student as well, you have to understand that if you don’t get a job right away, it’s nothing personal. It’s nothing personal. You move on and you keep trying, you keep trying. You have to keep trying.
Josh Hortaleza: 00:56:55
It’s something that I really resonate with because I was unable to break into the equity analysis space, right? And even though I shifted gears and I pivoted to this field, I still have an understanding of this notion of grittiness. Because especially when you’re a young kid, you believe that you’ve lived a long time, right? You believe that you know things, but no. You really have not.
Josh Hortaleza: 00:57:24
You’re just forming your career. You’re forming your career and you want to learn as much as possible and you have to really understand that okay, even though I can’t get a job right now, it’s not the end of the world. You have to keep going in a direction that eventually gets you a job. Whether that be within the form of self learning, whether that be in the form of getting an education, whether that be in the form of networking, whether that be in the form of building up your portfolio.
Josh Hortaleza: 00:57:51
These are things that you have to keep doing. And the reason why it’s important to be gritty is because getting rejected is the worst feeling. It’s horrible. We hate getting rejected. It can seem like when you’re getting rejected from these companies that hey, is this endeavor even worth pursuing? And what I say is you never know until you try, right? So, keep trying, keep being gritty and understand that even though you might not be in data science now, you could be in it some day, right?
Josh Hortaleza: 00:58:29
And of course, you have to keep evaluating what options are good and what options are bad and what not. But if you’re truly dead set on this field, on this new field of data science, I truly believe that you have to have the wherewithal to keep going at it. And if you are truly passionate, if you are very passionate about the field anyway, you would have no problem continuing to expand your horizons and learn more about the field.
Josh Hortaleza: 00:58:57
So, you would already be doing that. So, keep your head up. Be gritty and understand that even though you are not working in data science now, it can definitely happen for you. So, that’s what I would say.
Kirill Eremenko: 00:59:10
I love it, Josh. Totally love it. I feel the passion, I feel the energy. I’m sure our listeners feel it too. Man, thank you so much for coming. Before you go, where are the best places to find you, connect with you and for people to network with you?
Josh Hortaleza: 00:59:27
Yeah. My name is Josh Hortaleza on LinkedIn. So, Josh and then my last name, H-O-R-T-A-L-E-Z-A. You can reach me through LinkedIn primarily. And that’s probably the best channel to reach me at. And yeah, I think that’s how I’d recommend it.
Kirill Eremenko: 00:59:48
Awesome, awesome. We will include that in the show notes as well. And one final question, what’s a book you can recommend to our listeners?
Josh Hortaleza: 00:59:54
I recommend the book Deep Learning by Aaron Courville, Ian Goodfellow and Yoshua Bengio. The reason why I recommend this book is because this book helped me a lot at my internship at Intapp. It was really able to make me understand neural networks at a very high level as well as go into the grittier details or the more fine details about how to improve, how to select parameters and what not. And how to optimize different aspects of the neural network.
Josh Hortaleza: 01:00:26
And so, from a practitioner’s perspective, not to call myself a deep practitioner, but it definitely did help me out when I was stuck learning what deep learning actually was. And yeah, that book was incredibly helpful to me as a data scientist, for sure.
Kirill Eremenko: 01:00:46
Got you. It’s also available for free if anybody’s interested, DeepLearningBook.org. Or if you want to support the authors, you can buy it on Amazon. Just look for Deep Learning, Yoshua Bengio. Fantastic. Thanks very much Josh, it’s been a huge pleasure to have you on the show. Really appreciate your time and good luck with your epic journey. You’re going to go very cool places man.
Josh Hortaleza: 01:01:09
Thank you very much for having me. I really do appreciate this. And I hope that we can stay in contact.
Kirill Eremenko: 01:01:19
How amazing was that? How do you feel? I feel super pumped. We finished recording this podcast 30 minutes ago and I still feel the energy. I feel like I need to go for a run or something like that, I have so much inspiration. I hope you got as much value out of this. My personal favorite part was, and the most surprising part for me was that Josh at 22 years old is designing his career.
Kirill Eremenko: 01:01:46
Even his own internships by design of four months long. I’m sure he has opportunities to stay longer or look for full-time jobs and things like that but he by design … Just look at his LinkedIn, four months every single time. His internships by design are four months long. That is a true testament to a vision and a thorough follow through on that vision. Nothing is going to stop this person, he is going to get to where he is going.
Kirill Eremenko: 01:02:21
And it’s a great inspiring example. If you want to build a great career, design it and follow through on your vision, on your strategy. Amazing, loved it. As always, you can find the show notes at SuperDataScience.com/389. Connect with Josh, make sure to hit him up, network, ask him questions. I’m sure he’s going to be happy to answer them. If you’re in the space of fundamental machine learning research in Canada, in the US and the UK, wherever, in companies like Facebook AI, Google Brain, Uber, and so, hit Josh up.
Kirill Eremenko: 01:03:00
He would love to get in touch and get an opportunity to be in that space. And as you can see, he’s going to add a lot of value. He’s going to be the person. He’s a go getter. All right. And in addition to that, please share this episode. If you know somebody who is super excited about data science, doesn’t know how to break into the field, maybe is a bit scared, maybe is a bit unsure that they can do it, send them this episode. Super inspiring, super energized. You might change somebody’s life.
Kirill Eremenko: 01:03:28
Very easy to share, send them the link. SuperDataScience.com/389. And one final thing, DataScienceGO Virtual coming up this October. Number two. We’re going to have like last time, we had two and a half thousand people. This time we’re probably going to have, we’re aiming for four thousand plus. So, make sure to be there. It’s free. Just go to DataScienceGO.com/virtual. Sign up, sign your friends up, sign your whole family up.
Kirill Eremenko: 01:03:56
Of course, you need to fill in the survey. Sorry, you need to fill in an application form because we want to make sure the right people are there so we’re able to network with each other. So, fill in that form, get yourself there and meet amazing people like Josh and add them to your network. Now you have the tips, the ways to network. Read the book by Dale Carnegie, How to Win Friends and Influence People. Turn up, show up and take your career to the next level. DataScienceGO Virtual this October, can’t wait to see you there. And until next time, happy analyzing.
