Data Driven Futures: Adapting and Thriving in an AI World | Navigator Series

Note: This episode of the Navigator is better if heard. If you are able, we encourage you to listen to the audio as the text may contain errors due to the hard-to-follow nature of a complex discussion. The transcript was generated using a mix of speech recognition software and human editors, and can only be edited to a certain degree without losing the nature and the meaning of the conversation. Please check the audio before quoting in print.

Welcome to the Navigator Series presented by Lighthouse Labs. In this episode, the CEO of Lighthouse Labs, Jeremy Shaki sits down with two data leaders, Kishawna Peck from Womxn in Data Science and Ramnik Sandhu, a Senior Data Analyst, to break down misconceptions surrounding the data profession, how women can (and should) break into data, and how soft skills take you further than you think.

Huge thanks to Kishawna and Ramnik for taking the time to share their wealth of knowledge with us.


Data-Driven Futures: Adapting and Thriving in an AI World | Navigator Series

Or listen to the podcast.

00:00:00:00 - 00:00:26:05
Jeremy: Hi, my name is Jeremy Shaki, co-founder and CEO of Lighthouse Labs, and I'm so happy to welcome you to the Navigator Series: a series of panels featuring leaders in cybersecurity and data from across Canada discussing their tech journey, what's impacting the current job market, and what to expect from the future of work. Today's session is titled Data Driven Futures: Adapting and Thriving in an AI World.

00:00:26:11 - 00:00:58:05
And we're going to be discussing how we build data driven cultures. And what will AI’s role look like in the future of work. Ready? Let's get to it.

Jeremy: Hello. I'm joined by two amazing leaders in the Canadian tech scene. To my immediate right, Kishawna Peck from Womxn in Data Science, the founder and CEO of it, in fact. Welcome, Kishawna.

00:00:58:09 - 00:01:19:06
Kishawna: Thanks for having me.

Jeremy: And to her right, we have Ramnik Sandhu, a Senior Data Analyst. Welcome, Ramnik.

Ramnik: Hi, Jeremy. Happy to be here.

Jeremy: So happy to have you both here. You both ready to dig in?

Kishawna & Ramnik: Yeah.

Jeremy: Let's go. So why don't we start off by just understanding a little about each of your journeys? I feel like you both had some really interesting paths into the careers you're in.

00:01:19:08 - 00:01:43:22
Why don't we start with you Kishawna?

Kishawna: Okay, so I'm going to start at grade 11—

Jeremy: Nice, early.

Kishawna: I feel like that's where the shift happened. I was taking grade 11 math, and my teacher at the time was like, “You actually don't need to grade 12 math. You shouldn't do that because it's going to ruin your university chances”. So I didn't take Grade 12 math. So now I had to apply for university.

00:01:44:00 - 00:02:05:19
And a lot of the subjects required me to have math. So I literally just deduced what I could apply for and applied to that. So I ended up in sociology at the University of Ottawa, and I would do my papers the night before. They were too easy for me. The procrastinator. They were too easy for me.

00:02:05:20 - 00:02:29:19
And I said, I need a challenge. I want to do something challenging, which you'll see is a problem for me.

Jeremy: Okay, this is a repeat.

Kishawna: So I ended up switching to economics and I switched schools to York. I took economics. I had to teach myself grade 12 math before I could even do any of the courses.

00:02:29:21 - 00:02:52:20
Surprisingly, when switching, they didn't notice that I didn't have any prerequisites, which was nice, but I had to teach myself calculus. I failed stats twice. A C was like “You made it!” I used to always fail my midterms but passed my finals. And it wasn't until like fourth year that I was actually getting A's. So it's like, I kind of like these subjects.

00:02:52:22 - 00:03:19:17
And when I graduated with my economics degree, I was covering someone on a mat leave for a financial analyst. And one of the things that I was doing was checking different Excel reports. And I was able to find some discrepancies. And then when I finished doing that, covering the mat leave, I went back again to figure out what I want to do with my life, which is something we do every few years—

Jeremy: The repeat

Kishawna: The repeat.

00:03:19:19 - 00:03:34:16
And I was like, I want to be a consultant. Because in my mind, consultants help people with problems, and in order to be a good consultant, I would need to be able to back it up with evidence. And I was like, let me figure out what that is. And then I found out that that was data analytics.

00:03:34:21 - 00:03:56:12
So I took a post-grad in data analytics, and while I was doing that, I also did an internship as a data analyst at the same time. I used that internship to prove to them that they needed a data analyst because they didn't have one. It ended up working. So I got my first data job. I worked there for a few years and then I switched companies.

00:03:56:14 - 00:04:18:03
I was working in the product department with product data initially, and then the company I changed to was like I was servicing all the different departments, but mostly in the finance. And I was working there helping with projects, broader projects than I should have been working on to the point that they were like, we're going to promote you and you need a team now.

00:04:18:03 - 00:04:46:02
So then that's how I ended up growing a team there.

Jeremy: After how many years?

Kishawna: I was there for like six months.

Jeremy: Wow. Amazing.

Kishawna: Yeah. So I ended up growing a team there and like, I was the most senior data person there. And before I had gotten there, I had run my first conference. Which was an accident because I wanted to go to a conference and I found one for Women and Data, but there was nothing in Canada.

00:04:46:02 - 00:05:05:18
So I was like, someone should probably do that. And as the oldest daughter of Jamaican parents, I was like, I planned Thanksgiving and I planned Christmas. Like, what was it? It was a big deal. It'll be in the back room of a library. It'll be something small. I planned it in six weeks. There were more than 100 attendees.

00:05:05:18 - 00:05:22:11
I was able to find sponsors one of the weeks I was actually in Mexico. I confirmed the majority of my sponsors while I was in Mexico doing Mexico things. So it's not like I was there not having fun.

Jeremy: Did they know that you were in Mexico?

Kishawna: They did not.

Jeremy: Okay. All right. I was just wondering.

Kishawna: I used Zoom in the back, they didn't know.

00:05:22:13 - 00:05:43:20
And then what? During the first conference, I noticed that it wasn't as diverse as I'd hoped it to be. Yes, it was a conference for women, but there was more opportunity to ensure that more diverse women were there. So we ended up making AI and data literacy programs and just different avenues for more diverse women to get into the space.

00:05:43:22 - 00:06:03:14
And in 2021, in July, I actually quit my job and this is what I do full time now. So it's a beautiful accident. But I think I fell into everything because now I consult. Now I do some consulting, but it's not—

Jeremy: You went full circle all the way back to the consulting. And what do you consult on?

00:06:03:16 - 00:06:26:17
AI strategy and data and literacy programs.

Jeremy: Okay, that's an amazing story. I have quite a few questions for you. Let me go to Ramnik and then I'm going to come right back to that. So. Ramnik, how about your story?

Ramnik: Sure. I will also start at university. I studied economics, similar to Kishawna. I studied French, studio art, as well as philosophy.

00:06:26:19 - 00:06:49:09
And my first role out of university was strategic communications at U of T. And then from there I went into social media and then I was doing performance marketing, a lot of analytics, and I just remember watching my peers, you know, pursue their next role in marketing. And I had a voice in my head like, why don't I want to do that?

00:06:49:11 - 00:07:17:07
I was always looking for a challenge, but for me, marketing was not challenging enough. So cut to 2020. Everybody was doing layoffs because of COVID. I was affected and I found myself with a lot of time on my hands. So what I did was I created myself a curriculum of sorts. I spent, I want to say like four months, day in and day out, full days just studying.

00:07:17:09 - 00:07:49:22
And that led me to my role at Walmart. I joined the merchandising department supporting analytics.

Jeremy: And so when you went into that merchandising role, did you deal with any imposter syndrome having just switched careers and changed places?

Ramnik: Oh yeah. The first month was really, really hard. I didn't know anything about merchandizing, and the realm of merchandising that I was dealing with was, you know, identifying KPIs in terms of how productive a shelf space is.

00:07:50:00 - 00:08:16:04
So I was responsible for making up those KPIs, measuring them across time, and getting other teams onboarded to that. And then just kind of standardizing the reporting process, stuff like that. But just to do that, I didn't have the foundational knowledge of merchandising. I didn't know the lingo. So the first month you know, I was close to a few breakdowns, definitely had imposter syndrome.

00:08:16:06 - 00:08:38:19
But I found that just arming myself with the knowledge of learning, I guess, like reading documentation, learning how Walmart's data sets it, and then talking to stakeholders. It just really helped me to kind of overcome that.

Jeremy: So some of the other skills that kind of came into play while learning the actual core skill in the data side.

00:08:38:23 - 00:09:01:11
Yeah, interesting. And your journey, Kishawna, you I mean anyone listening, probably there's like 11 or 12 instances where they could feel imposter syndrome talking through. Did you ever go through anything like that? Did you ever feel that at any point?

Kishawna: I feel like it missed me because I tend to move too quickly. So this kind of imposter syndrome misses me.

00:09:01:16 - 00:09:22:21
However, there were different points of discouragement, like external discouragement that I've received. And I would say my biggest, Boogie Man in the closet would have math. But I've seen him so many times now that I think we're good.

Jeremy: You can look at each other and be okay?

Kishawna: Yeah.

Jeremy: All right.

00:09:22:22 - 00:09:50:06
And so if that's the case, like, what are the skills? I guess that if it's not math, which everyone would assume is so core with data, what do you think the skills are that kind of propelled you forward?

Kishawna: I would say problem solving and resourcefulness, which were core things that I got from my economics degree. Like with the economics background, those optimization problems they give you, you have to be resourceful, you have to problem solve.

00:09:50:08 - 00:10:06:22
And I found even when I didn't know things that I was dealing with at work, I was resourceful enough to figure it out.

Jeremy: I mean, even like you set up the event, you set up and the way you did it is you really saw a problem, you just decided to solve it through using resourcefulness while in Mexico.

00:10:07:03 - 00:10:21:23
Kishawna: Yeah. Even the way that I set up the company I got my first sponsor and, and they were like, who do we send the cheque to? And I was like, me, because I don't have a bank account because I wasn't a registered business. Okay, so then once they said that, I was like, I need to incorporate.

00:10:22:04 - 00:10:40:13
So then I incorporated them and the next week they were able to send me a cheque, but I didn't even know that that was a thing. So I think a willingness to learn is also a big skill in data too. So just being open to figuring it out as you go.

Ramnik: Does that resonate for you, Ramnik? Like where your skills, you know, is it math?

00:10:40:13 - 00:11:00:08
Like what were the core essential skills for you coming in?

Ramnik: I've always been good at math, so I can't say that I really struggled with it. But I think one of the core skills for me has been more the soft skills. I think communication and critical thinking and working with other teams and how to best collaborate with them.

00:11:00:10 - 00:11:29:19
I think that's been really important for me to not just like problem solving, as you said, to design the solutions, but also I think the collaboration to make sure that teams are using those solutions. I think that's been a really, really pivotal step I think in my journey.

Jeremy: Amazing. And do you think, you know, people ask this a lot at Lighthouse, the difference between data analyst and data scientist and the skills needed.

00:11:29:21 - 00:11:49:12
Do you see a big difference in those skills necessary?

Ramnik: Not so much. I think if you have the right foundations. So in my journey, what I did, one of the things that I did was I created an audit essentially of all the job postings that I was seeing at the time in March 2020 - there were quite a bit of postings.

00:11:49:14 - 00:12:11:11
And I compared analysts as well as data scientists, and it was all the same lingo across both. But even in practice, in my role, like I do a little bit of predictive work, I do a lot of prescriptive work, but I do think there's a lot of overlap. There's a lot of overlap and I don't think that you necessarily have to be an analyst first and then be a data scientist.

00:12:11:12 - 00:12:36:05
I think that if you have the right foundation, you know, like the right machine learning principles, if you arm yourself with, you know, you learn Python and SQL, I think those are like foundations. And you can go straight into being a data scientist.

Jeremy: Okay, So I mean, that brings me to, you know, you have these fields and they're kind of growing spaces, and we talk about the idea of AI.

00:12:36:05 - 00:12:54:07
And the first thing at the core of AI is like, well, data, right? There should be a lot more data jobs and there should be a lot more people jumping into data. Yet, and, you know, listening to both of your stories, they're both stories of people who kind of just entered in. You called it a magical mistake.

00:12:54:07 - 00:13:14:06
A beautiful mistake? A beautiful mistake. I like that. So, you know, like a lot of people's careers end up in that beautiful mistake space. And you didn't seem very intimidated by entering in. And you said you, you know, you outran imposter syndrome. So how do you feel when looking at the data space and seeing not nearly enough people entering in?

00:13:14:10 - 00:13:42:03
Why do you think that is? I may be inferring a little bit too much here, but considering what you're running Womxn in Data, what do you see as those barriers for people coming into the space?

Kishawna: So there's I think it's like approximately 23% of data roles are held by women. I think people, especially with the AI boom that's going on right now, a lot of people think the train has already left when a lot of companies don't even have proper data infrastructure in the first place.

00:13:42:05 - 00:14:22:04
Jeremy: Let's fix that.

Kishawna: So you're not late. Also, people think you need to have the title of Data Analyst, Data Scientist, Data Engineer. Meanwhile, there's other roles like Product Manager that can use the same skills, or even working in marketing, there's roles where you're still using data - that being someone who has data skills that complements it. So in my work, I'm actually seeing more people that are looking to transition or complement their main skillset with data skills and looking at it that way is probably a better way of looking at it, and also feeling like you need to outrun what is happening in the AI space.

00:14:22:06 - 00:14:49:07
So being like, okay, like what is like a bullet proof job I can get? I feel like if you focus on the foundational skills that we've already been talking about, like a willingness to learn, resourcefulness, and problem solving, that can be applied to so many different roles. Like it doesn't need to have a certain title.

Jeremy: Yeah, I mean, I'd say at Lighthouse Labs, like that's where we're training for more is definitely the problem solving.

00:14:49:07 - 00:15:12:18
Learning to learn at a good pace, to be comfortable being uncomfortable, and how you do that. But do you think that the barrier for many people coming in— first of all, we can talk about external barriers, but just even from internal and who people are, do you think this is more of a data literacy issue or is it like people not even getting to data literacy?

00:15:12:20 - 00:15:36:22
Do you think it's more about the jobs of data analyst and data scientist? Like which way would you look at that?

Kishawna: I think it's more about data literacy. I think the average person doesn't realize that they're already interacting with data power tools every day. And if they were able to understand that they're already interacting with those tools, then it would be easier to want to take the veil off of what's going on behind the tool.

00:15:37:01 - 00:15:56:19
Jeremy: Right. And Ramnik, when you were entering into the field, did you feel like there was any like internal, you know, we talked about the imposter syndrome of like the merchandizing side, but, you know, actually, like when it came to the actual data and the math, you said you're comfortable with it. Was there any internal barrier for you and was there any external barrier that kind of came into play?

00:15:56:21 - 00:16:22:02
Ramnik: Yeah, I think both. Like internally, I think, you know, I definitely felt a certain type of way just seeing everybody, you know, continue their linear career paths. And I was like, I want to do something different. So that was more so internal. But prior to me actually, you know, taking that first step of creating myself a curriculum, I had actually wanted to make a career change.

00:16:22:02 - 00:16:44:03
I want to say for at least like a year and a half before the layoffs happened. And I just remember I had a conversation with a former colleague of mine, we used to sit right next to each other in the office, and I mentioned in passing that I really wanted to, you know, learn about data science and, you know, like I was the response that I received was very much, you cannot do it.

00:16:44:08 - 00:17:08:14
You have to go to school for four years. You have to do a computer science degree. And that really, you know, shut my hopes down for this, for I want to say at least like until I was laid off. So in terms of like I think there's a lot of intimidation in that sense. And I think people also think that it's much harder than it actually is.

00:17:08:16 - 00:17:27:20
Like you said, I think you have to have a willingness to learn. If you have that, I think you can make that career change.

Jeremy: Does that resonate for you and what you're seeing in Womxn in Data like the person suggesting that it's not for Ramnik, like, how are you finding the feedback from all the people coming around?

00:17:27:20 - 00:17:53:16
What has been there that external barrier?

Kishawna: I feel like a part of a part of the work I do is even making people realize that they already work with data in their in their job, especially if they have like a research background or like even something they took in school required them to do any form of research, like just making them understand that they're already doing parts of the role.

00:17:53:18 - 00:18:16:13
And even, like I just finished my master's of management in AI and I was so scared for the AI math course. Like, I cannot explain it to you, Like I was so scared. And then when we started the course, I was like, this is grade 12 math. This is grade 12 math that they wrap up, and it makes you think that it's difficult, like it's linear algebra, it's stats, it's calculus.

00:18:16:15 - 00:18:47:12
So if there was more emphasis on taking away that math phobia from the beginning, then it wouldn't seem like it's this big thing to engage with the math. And then there's also like, it's not like we're sitting down figuring out functions, you know, and like, explaining them, like we're not we're not doing that. So there's the level that you need to understand of the math isn't even as deep as someone not in the space thinks that it is.

00:18:47:13 - 00:19:10:02
Jeremy: Okay, so then what is it like? Because that's, you know, I think that I would agree. I see so many people who really feel it's when you mention the word data, data analyst, data scientist, data literacy. It's math, right? It's functions— what is that? What are these things? What are people needing to do when we talk about data literacy or beyond?

00:19:10:04 - 00:19:33:21
What are people doing with it?

Kishawna: I think there's definitely a necessity to understand the math behind it so that you can check your work. And so that you understand the different models that you're using, things like that. But I don't think you're necessarily like providing the proofs for a function.

Ramnik: Yeah.

Kishawna: Which I think that's the type of things that make people be like, I don't want to do that.

00:19:33:23 - 00:19:54:08
So I feel like it's more, it's more problem solving, it's more framing the question, making sure you're actually asking the question that they want answered. Because half the time the business doesn't even know what they really want. Detective work of you trying to be like, okay, let me put this together. So that's where I'm getting domain knowledge on the job. It is really good.

00:19:54:08 - 00:20:23:11
You don't necessarily need to with data. You could be in different industries and just learn as you go on the domain side. So I would say like 50% of it is the problem definition. And then anything after that, like from collection to finishing it is basically ensuring that you have the right mathematical parameters for different things that you're pulling, like making sure when you pull the data, like what does the sample look like and the information like that.

00:20:23:13 - 00:20:45:07
So that's where the math goes in. But it's not as deep. It's not as deep as I would have thought it was. Because even with all my experience, when last year I picked up that book, I was like, this is not going to go well. When I opened it, I was like, wow, we're really being bamboozled.

00:20:45:08 - 00:21:00:23
Jeremy: Be fooled. So I saw you nodding your head. So I'm going to assume you agree with everything said here.

Ramnik. Completely. I use the example of linear regression. It's literally just Y=mx+b, and we learned that in grade ten.

Jeremy: I mean, you just said that real fast. That scared me a little bit.

00:21:00:23 - 00:21:24:16
Ramnik: Sorry.

Jeremy: Okay. Go ahead.

Kishawna: It’s the equation of a line.

Ramnik: Yeah. It's literally like we learned that in grade ten, right? So it's not as intimidating as people make it out to be. I think she said it perfectly.

Jeremy: Do you then, like when you're evaluating a junior person coming in, okay, and they want to get into that first data job and, you know, some people have learned these skills or tried to learn.

00:21:24:16 - 00:21:46:05
There's a lot of free content, you know, good plug for Lighthouse Labs, but someone comes into a junior job. What matters, like you said, the resourcefulness and problem solving. But how are you actually evaluating their skills? Are you going more on the data lexicon, the skills in data itself, or are you going more into the essential skills of what people like just have in them?

00:21:46:05 - 00:22:05:15
And believing in that plays out well for data, like, how are you doing that?

Ramnik: For me, I want to see that a junior data person is curious. Are they asking the right questions? So I think in my experiences so far, I think you mentioned having domain knowledge? I really needed to do that myself.

00:22:05:16 - 00:22:27:02
But I'm not I'm not talking about myself as an example. But I think just being willing to ask the right questions and understand what the problem is that our business teams might be facing. I think that is one of the things that I look for. Just because you're not going to be able to problem solve if you don't understand what the problem is

Jeremy: Love that. Yourself?

00:22:27:05 - 00:22:49:21
Kishawna: For me, it would be if they are not even curious about things like to add on, it would be like questioning themselves, like are you going to check your work ten times before you bring it to me? Because at the beginning when I was onboarding new staff, it would be like I could look at something for 2 minutes and I could tell that you didn't do X, Y, and Z.

00:22:49:23 - 00:23:20:14
So I trained my team to basically you need to anticipate the questions that are going to come from the work that you're doing. So you already have an answer for that— hence how this panel is going— and just like that sense of over preparation and asking the next questions. So being curious about the work that they're doing. Technical skills, I used to have an education budget and we can fill in whatever technical gaps you have. And in terms of communication skills and soft skills,

00:23:20:14 - 00:23:46:15
I was more putting them in situations that they would be able to refine, their communication skills, their collaboration skills and things like that. So I feel like I'm definitely fine with molding people into the potential of what they could be. And I more want to see, like I didn't care about credentials, like where you went to school, anything like that.

00:23:46:15 - 00:24:08:14
It was more, can you explain to me, do you have a project that you've done before and can you explain it to me? That's how I took interviews.

Jeremy: Yeah, we hear that. We hear that a lot. And I think that sense of what the person needs to do to show their success is not what a lot of junior people think that they need to show

00:24:08:14 - 00:24:27:11
coming into job interviews. Coming into kind of, you know, they feel like that certificate, that skill set, what they can say is a learning opportunity. Whereas you're actually looking at how diligently committed are you going to be in the problem solving function, right? Like how hard are you going to drive at this and how can I rely on you to care about it to that level?

00:24:27:13 - 00:24:48:16
We hear a lot of people and by the way, I heard the self props— they're saying, you know the prep for this panel, I'll I agree it is going excellent. So with the idea of AI jobs. There's this bigger discussion coming up and it's like, you know, we see it. We see the rise in people saying, I want to get a job in AI.

00:24:48:18 - 00:25:12:11
What does that mean for you? What are jobs in AI to you and what would you tell somebody who says that to you.

Kishawna: I would tell them to look at the AI pipeline and decide along which point they want to be on it. So I feel like there's a lot of different roles that go into it, and it's not necessarily one role.

00:25:12:11 - 00:25:35:21
Like it's not like an AI engineer, it's not a deep learning engineer. Like there's a lot of different roles that go into it. So I would actually— and also, not even roles, like do you want to build something? Is there something you have an idea to build that's AI related? I feel like it's very— I would ask them what they think AI is. That’s where I’d probably start.

00:25:36:00 - 00:26:10:03
Jeremy: Okay. Do you feel like data learning in the data space is kind of a core starting point for moving into most roles? Or do you think there's a wider gamut right now that is possible.

Kishawna: You definitely need foundation or else you could be making mistakes everywhere, you're not going to understand like the supply chain of the data that you use, which is very important, especially with more talks on like responsible AI, because you, you want to make sure that you understand ethically what you're doing.

00:26:10:05 - 00:26:35:02
So I would say it's foundational. Maybe you don't necessarily need to go down a specific path, like maybe you don't need to take something that's specifically for this type of role. But having a foundation of data literacy is a requirement because like AI is a tool, but data is the raw material. So how you can use the tool and you don't have any with what do you understand what to do with the material?

Jeremy: What a great statement.

00:26:35:02 - 00:26:51:16
Yeah, I totally hear that. And I do want to come back to responsible development because I know you are pretty passionate about that. Before I go there, I guess. Ramnik, doing the role of a Senior Data Analyst and you hear all this stuff around AI: it's walking in the door. What does it feel like in your type of role?

00:26:51:16 - 00:27:17:02
Like, what are you thinking about for yourself when it comes to AI?

Ramnik: I think when it comes to AI—so I still consider myself early in my career— but when it comes to AI, I use it to help me be more efficient. But I do think that if I didn't have the foundational knowledge that I do have, I don't think it would be as useful as people are making it out to be.

00:27:17:04 - 00:27:37:19
And then and the other thing with that being I've worked for a really large company being Walmart as well as a smaller company. And what I've seen also goes back to data literacy. Teams are not yet ready to rely only on AI. I think there's a lot of groundwork that needs to be done before we can get to that.

00:27:37:21 - 00:27:55:18
Jeremy: Do you mean that just like companies are too rudimentary in how they're using data for AI to really make a difference? Is that what you're saying?

Ramnik: Yeah. Companies are very rudimentary, but also I think there's like a, like a skills gap in all of these companies, all of the companies that I've been in at least.

00:27:55:20 - 00:28:33:07
And it goes down to people being afraid of big data. I know we've seen big data being thrown around a few years ago. People are still intimidated by that term. So I think that kind of plays into that.

Jeremy: I think that's a yeah, I feel like most of the companies I talk to, it's very surprising how light they are on data skills within non-data jobs. Like you know just across the org considering what the warnings were for years and all the big data is coming in, you know data transformation, we have so much at our disposal, and yet it really feels like there's been there's still a massive gap in understanding

00:28:33:07 - 00:29:00:00
what to do with it, let alone kind of moving on to A.I. You agree with that?

Kishawna: Yeah, I think that on AI maturity, the majority of companies are like a 1. Like, they're even— a big problem a lot of companies do is like they'll bring in data scientists and data analysts and do not have proper data infrastructure, like they don't have the proper pipelines to bring in their data, store their data.

00:29:00:00 - 00:29:26:10
Like those are basic things. So I know now, like everyone's speaking about AI, but a lot of companies still have those same data, like infrastructure problems from before. So how are they supposed to go to that? How are they supposed to skip to the next step, which of course there are tools, but for a lot of small and medium companies, those tools are like not accessible for them financially.

00:29:26:12 - 00:29:54:06
Jeremy: So how do you feel? I know you've done work on the data science side. You're looking at a lot of events across Canada right now, that's kind of talking about AI and what's going on. How do you feel from an ethical portion, like this stuff is impacting industry and we're saying it's not quite there yet. What do you think about responsible development of AI right now?

Kishawna: For responsible development?

00:29:54:06 - 00:30:23:08
What I'd like to see more is companies following a responsible AI framework. Also, just because there's so many big players right now and the lagging of policy, I feel like it's not an optimal mix. And I feel like the general public is also not engaged as much with this, especially since you're worried about your jobs.

00:30:23:11 - 00:30:47:12
Yeah, so there's like this we're in a very interesting time right now for responsible AI. And like one of the reasons why I took my Masters of Management and AI's, because I heard a lot about responsible AI and like I spoke about it a lot myself, but understanding the actual interventions and the product development cycle was something that I wanted to dig into.

00:30:47:14 - 00:31:13:03
So looking at it from ideation, all the way to monitoring after deployment. So I feel taking that type of lens and seeing what are the actual interventions companies can take as they're developing is really important. Because if you just take it as like, “yeah, we're ethical”, but you're not doing anything in the different stages. It's just a statement on the website.

00:31:13:08 - 00:31:31:09
Jeremy: I mean, and that can be, that can be responsible “a lot of things” if you're not actually going deep into it and thinking about it properly, it just becomes a marketing statement. Right? What so to you then, what is responsible AI? Like, you know, you went and did your master's in this. Like I think most people, they hear those words,

00:31:31:09 - 00:32:09:10
they're not even sure, as you said, what is AI to you, period? Let alone what is responsible AI? So what does that mean?

Kishawna: I would say responsible AI is an AI that has thought through things like where their data is sourced, how it's being used, the privacy of the people that it's being used with or on. Consent, just different like different aspects of— I feel like I'm falling into merchandizing— of the supply chain of whatever product or service that you're putting out.

00:32:09:12 - 00:32:32:00
So from ideation, are you working with different minoritized groups to understand how this product would impact them? And data collection, are there certain groups that you are counting and who's not being counted, who's not being counted as a really strong one? Because sometimes data excludes different communities just because they don't even have— they only have like a dropdown for them.

00:32:32:03 - 00:32:56:20
So it excludes them also, like during development, like which models are you using? And like do those models also exclude or like over-represent different groups in deployment? Like where are you deploying? Are different groups at a disadvantage now because of it? And then what kind of governance framework do you have in place to catch your mistakes before It's a headline?

00:32:57:00 - 00:33:25:14
Yeah, because you know, no one wants to be a headline. And also like, how do you fix those things? Like what's the actual plan once it pops up. Like one of the slides in one of the talks I have, it's full of different headlines that are AI related. So in terms of child welfare, in terms of medical mishaps with women, in terms of not being able to get a mortgage because the AI decided whether or not you did.

00:33:25:17 - 00:33:48:08
Yeah, just different aspects. And I feel like once it starts impacting your different needs, remembering Maslow's Hierarchy of Needs, like once it starts impacting like those two bases at the bottom, it's kind of like, “What are you doing?”

Jeremy: Yeah.

Kishawna: But there's also a different innovations that if it hits like the love part. So like there's algorithms for dating, right?

00:33:48:13 - 00:34:11:20
Like, are you blocking me from love? Like knowing me like, that's a more like, frivolous situation compared to the other bottom parts of Maslow's Hierarchy of Needs.

Jeremy: Are you, are you blocking my mortgage? Are you blocking my mortgage from getting a job?

Kishawna: Exactly. Is this a safety issue as well, right? So I feel like it's not just those fundamental things, but those are really important.

00:34:11:20 - 00:34:36:18
But just like if you touch any of that, or if you even have to question if you're going to touch any of that, I would definitely take a closer look at your entire product development cycle.

Jeremy: So, I mean, great answer. And, Ramnik, you said, you know, you're playing around with some AI. Do you find yourself in a community filled with data analysts,

00:34:36:20 - 00:35:00:02
is this a conversation that's regularly coming up, the idea of responsible AI? Like, do you see it around you?

Ramnik: So with other data analysts that I know, we kind of laugh at how often ChatGPT doesn't give us the answer that we want. So we're often like, you know, you have to be very, very specific with what you ask ChatGPT just to get the answer that you want.

00:35:00:04 - 00:35:22:23
And then you have to know what answer you want, right? So I think that's kind of the running joke that we have—

Jeremy: Obviously, and I've heard this a lot, AI is not quite at the level that it's doing what people want it to do, right? But then you have this other side of the ChatGPT or not, I'm making certain decisions that can really impact people's lives.

00:35:23:01 - 00:35:38:00
And at the heart of data, what you'd want to know, like as an outsider to data, what I'd want to know is, okay, are all the data people very aware that how they're building and how they're shaping in the kind of stuff they're feeding, how it impacts AI But is that even possible? Like how, how do you—

00:35:38:00 - 00:36:07:19
Ramnik: Yeah, I think like being early in my data career, this is the type of topic that I'm only now starting to think about and I've seen discussions about how AI is impacting other fields. So we are very cognizant of, you know, AI in that sense. But I think like for me specifically, I think purely because I'm still very new, it hasn't been too deep of discussions.

00:36:07:20 - 00:36:28:01
Jeremy: Yeah, okay, that makes sense. And I think that's where we are, you know, with all this transformation. And back to the point of the big data, what it's supposed to promise, like we're still so early in so many things that then you get into kind of some people moving way with certain AI algorithms, certain pieces that are just like limiting people significantly.

00:36:28:01 - 00:36:47:10
And then there's a whole bunch of the rest of everybody who's like just trying to catch up on the starting points of this. It feels very broken. It feels very separated when you're looking at organizations and what they can do to do better with data, like, you know, we hear we've heard of data-driven organizations.

00:36:47:12 - 00:37:19:04
What is the organization need to do to be more prepared to be a data-driven organization, especially in this world with this stuff coming.

Kishawna: So I think companies should actually be more data-informed than data-driven, and that's to give more credit to people: their instincts, their like backgrounds, like things they know. Because I feel like if you're completely data-driven, you miss a lot of things that aren't coded in the numbers.

00:37:19:06 - 00:37:38:06
So I would say for a company to be more data-informed, going to go back to infrastructure, having the right infrastructure, ensuring that you're collecting data that you need, and a lot of companies have just like been collecting data and just don't know what to what to do with it, because they never had a plan for it in the first place.

00:37:38:07 - 00:38:04:15
So kind of looking at what you have and deciding what can come of it, which is more like in the data scientist realm. And from a responsible lens which anyone can do, you do not need to be in data, is to think about how the decisions you're making or the tool that you're building in your company impacts other people's actual lives.

00:38:04:17 - 00:38:23:18
Jeremy: So I think about the person and the customer. Yeah, you know, you're supposed to do that. It's easy to do that. It's crazy.

Kishawna: It's crazy. As they always say. It's supposed to go like, you got to be user-centric, right? Like that's a study that has nothing to do with data,

00:38:23:20 - 00:38:50:21
but lends itself well to the field as well, right? So someone who's looking to transition or complement their skills and their like in user research or something like that, this is perfect for them. So understanding who you're building for and how it impacts them is the core. Like if you run with that, you're mostly good because then you need to figure out how to actually implement things to protect the people that you're building for.

00:38:51:01 - 00:39:12:14
Jeremy: Yeah, I think when you say the difference between the data-driven culture and the data-informed culture, you're really putting people at the top when you talk about a data-informed culture. And your point, I mean, all your points today have been very driven towards the people, right? Data is here to help people, not people are here to help data just go do its thing and end up costing us.

00:39:12:16 - 00:39:31:11
Okay. So, Ramnik you know, when it comes to the idea of data literacy, obviously, we want people to gain it so that it's better for organizations. And we're trying to put people first. And by putting people first, more people understand the data itself is probably very helpful for all those different roles and jobs that actually do stare at customers.

00:39:31:16 - 00:39:54:05
But you're in the kind of job where you probably, I'm going to guess, interact with a lot of people who are quite reluctant to learn and listen to data and just try and throw it at you. How are you dealing with that problem and how do you get people to be open to learning and listening to data and learning more without making them feel like the data is taking control?

00:39:54:07 - 00:40:25:08
Ramnik: Yeah, that's a great question that is definitely a challenge in both of my roles so far. One of the things that I like to do is empower them to find the same numbers that I am giving them. So I do a lot of training sessions, so that's very hands-on and it's time consuming. But the idea there is to empower them to find the data themselves and I'm always open to, you know, it's the math.

00:40:25:08 - 00:40:48:16
People are afraid of math. I'm not afraid to, you know, like to explain math to them. Like sometimes the curious ones, they really do want to know the math. So I think just greater collaboration. And then one of the other things that I did at Walmart was to foster that data-driven, data-literacy culture I created

00:40:48:16 - 00:41:12:19
essentially, it was like a club where I was teaching them foundational data principles. You know, relational databases, what is an aggregation, what is a dimension, just very basic things and just show them that it's really not that complicated or that difficult. And I think the greatest joy for me is just like when I can see people have improved.

00:41:12:21 - 00:41:41:14
So I think just focusing on the collaboration and just understanding that, you know, what is simple to me is not going to be very simple to other people. I think that's been a big learning for me.

Kishawna: Yeah, we definitely have that in common because at my last company, because my data team at first was really small, I negotiated with the leadership team to train 1 to 2 people per department on different data skills.

00:41:41:14 - 00:42:07:13
So it was like a few weeks and it was over lunch. And like that also helped them to just understand that they could find their own numbers, and also reduce like the ad hoc requests I was getting. So that was helpful.

Jeremy: I mean, it's self serving, right? Like, you know, not even saying it in the wrong way. But it is self-serving in the sense that if you want that adoption, you kind of need the buy in and the buy in comes from people being open minded enough.

00:42:07:13 - 00:42:28:11
And so you go and it sounds like both of you have gone out championing that open mindedness, which, I mean, you know, you've done not just internally in your organization, but you've done outbound as well and through all your events, which is quite amazing. Like I think it does, it does feel just listening to both of you, like the need for curiosity.

00:42:28:11 - 00:42:52:22
You know, we see this in the cyber space just as much as the data space. It's domain knowledge, it's understanding business in general. It's being curious, it's being resourceful, it's being problem-solving oriented. That stuff is very normal business, and yet you stick a term on it like data or cyber, and it immediately feels kind of scary to a lot of people.

00:42:52:22 - 00:43:13:13
And you two have obviously been doing a lot of work opening that up and trying to make people feel more comfortable with it. Do you feel like you're being successful? Do you feel like you're getting people's minds opened on a regular basis?

Kishawna: I would say yes, but there's still a lot of work to do, especially since our goal is to inspire a million women to become data literate.

00:43:13:18 - 00:43:44:00
Like, that's not something you can do without partnership, without scale. But we keep coming up with better ways for more people, for education to become more accessible to more people.

Jeremy: Would you have three recommendations of things that matter most when introducing this stuff to people?

Kishawna: I would say, well, the first thing would be finding a problem that they care about and they want information for, so making it more accessible to them.

00:43:44:02 - 00:44:07:19
So for example, if you had a friend who, I don't know, has a horrible dating life, let's look at some of the features of your dating life. Like who are these people? What are some characteristics? You could service something like that or if they're really interested in, I don't know, knitting, choose a dataset that you're interested in exploring.

00:44:07:21 - 00:44:33:01
And then I would say take stock of what their skills are like, their innate skills are, and see how that complements the work that you're doing. So if someone is already naturally curious or they're naturally a problem solver or they're really good at talking. Like a lot of people that are really successful in data that you see in the more senior roles, they can cross both business and technical teams.

00:44:33:03 - 00:44:53:04
They are true translations.

Jeremy: Yeah.

Kishawna: Like that's actually a rule. So you don't need to be like, sometimes they’re a master of none, but they understand both worlds. So if that's who you are, why are you pushing yourself trying to be an AI Engineer when really what you need to be is an AI translator or product manager?

00:44:53:05 - 00:45:11:23
Not even so, just finding the right fit for you because it doesn't need to be. It doesn't need to be a specific box and half the time you can make your own box.

Jeremy: I mean, it is, right? There's I mean, both of you kind of did, right? Like you make your own space. You make your own. People are so in need they'll put you in that spot in the first place.

00:45:12:01 - 00:45:30:01
Right. Okay. So I know I know we're going to be wrapping up soon here. And so let me do a little bit of rapid fire for both of you, okay? 'm just here to scare you. Okay. We're going to go really quick, but we can look at a single word just like you know what that is?

00:45:30:01 - 00:45:53:07
What matters to you on that answer. So first of all, if you were to give advice to someone who's a little bit nervous about getting into data at all, what would you tell them?

Ramnik: Don't listen to other people?

Jeremy: Nice. My six words, we’ll continue. Go ahead.

Kishawna: Every day data.

Jeremy: Every day data. Okay. So not a rule-following group, but really great advice.

00:45:53:09 - 00:46:18:01
Okay. If you were talking to someone who's gotten into data and they're thinking about a true career in data, what would you tell them?

Kishawna: I would tell them to make a list of companies that they would want to dig into their data and find— are they supposed to be 1-word again?

Jeremy: Yeah, but that's okay. I'm letting it happen.

00:46:18:03 - 00:46:44:10
Kishawna: And find some data related to their industry and create a portfolio based off of that.

Jeremy: Okay. That was about 27 words.

Kishawna: I thought it was 46.

Jeremy: I'm doing the number thing. Yeah, that's not my strength either. So, Ramnik?

Ramnik: I would say be open minded. You know, once you've gotten your foundations, you really just need to get your foot in the door and it's not always going to be, you know, like the traditional job.

00:46:44:10 - 00:47:07:01
So I think just be open minded.

Jeremy: Okay, I'll come back to you if you were talking to an employer about how to encourage data literacy across the organization or an employer, try to encourage them to be a more data-informed organization, what steps should they take?

Ramnik: I'm going to go to what she was talking about, data infrastructure. Having the right leaders for data.

00:47:07:01 - 00:47:28:20
We were talking about this earlier, just having somebody in charge of data. I think that's step one.

Jeremy: Okay A leader driver. Okay.

Kishawna: So I would say understanding what problems you want to solve because then that will lead to what kind of data you would have in your data infrastructure and what kind of problems the data leader would be dealing with.

00:47:28:22 - 00:47:54:02
Jeremy: And the final one is, you know, we're at this point in time where people coming into this field can sometimes feel like they're not getting the role fast enough or, you know, they have to climb through hurdles to get to some pinnacle point. In three years from now, do you see a world where there are less data jobs and less need for data professionals than you do today?

00:47:54:04 - 00:48:17:07
Kishawna: No. However, I do see the names of the roles maybe changing and the data skills embedding themselves in other roles that you wouldn't normally traditionally see data skills there. So it's like a timeless skill. It's not going to disappear, but it might manifest in different ways.

Jeremy: Well see, this is why I tried to do the one word answer, because when you give me the whole one now, I want to ask more questions.

00:48:17:09 - 00:48:39:19
So just really quickly, give me a couple of those skills that you think change over the next few years. Maybe give us the predictive crystal ball.

Kishawna: I think maybe you won't need to code as much, because there's AI that can help you code. I think maybe some of the more technical skills, like maybe choosing which model fits your data best and things like that.

00:48:39:19 - 00:49:00:08
I've already seen tools like that. So maybe those types of skills you won't need. But skills such as reasoning, curiosity, willingness to learn, you need willingness to learn because you’re gonna need to learn because it's going to change. So I feel like it's a dynamic field and it's definitely different than when I got into it. So I'm not expecting it to be the same three years from now.

00:49:00:10 - 00:49:23:04
Jeremy: But okay, So question to you, do you see a world or there's less data jobs?

Ramnik: Absolutely not. Similar to her, I think the focus will be on other skills. Some of the more human skills, like she mentioned, curiosity, I think creative and critical thinkers are going to be really important. Just to bridge that gap between, you know, the data and the actual problems.

00:49:23:06 - 00:49:58:07
Jeremy: Awesome. Well, listen, both of you, I feel like this has been phenomenal. You both do such a great job of explaining larger, maybe intimidating concepts and making them very easy for people out there to go, you know what? That actually is something I could potentially do. You've, Kishawna, obviously dedicated a lot of time to removing intimidation for people while being a person that's very serious about the actual data side and not just kind of making it that message on a website that's, you know, “make this accessible”.

00:49:58:09 - 00:50:49:02
I guess I'll end off by asking you, is data for everybody?

Kishawna: Absolutely.

Ramnik: Definitely.

Jeremy: Awesome. And with that, thank you both so much for giving us your time for spending time with me. I've learned a lot. I hope everybody watching has learned a lot. And I can't wait to talk to you again very soon. Thank you, both.

Ramnik: Thank you, Jeremy.