Kesha Williams: Teaching & Mentoring in Tech, Machine Learning on AWS, and Balancing All the Things
Kesha joins Adam to discuss the lessons she's learned as a teacher and mentor, her favorite applications of machine learning in the cloud, and how she manages to balance her many passions.
Kesha Williams is an award-winning software engineer and technology leader that teaches others how to transform their lives (and livelihood) through technology. She's also an AWS Machine Learning Hero & Alexa Champion.
Adam: Hey, everyone. Welcome to AWS FM, a live audio show with guests from around the AWS community. I'm your host, Adam Elmore. And today I'm joined by Kesha Williams. Hi, Kesha.
Kesha: Hi, Adam. How are you?
Adam: I'm doing all right. I told you before the call, you're just a huge inspiration to me and if anyone discovers you because of this podcast or this Twitter space, then this whole thing has been worth it for me. If I shut it down next week and I sort of cancel all the rest of them. So I've just been really excited to get on with you. You've got a really interesting career and you've been in tech for I think 25 years. Is that accurate?
Kesha: It's now 26, 26 years.
Adam: Wow. 26 years.
Kesha: Yeah. I can't believe it.
Adam: You started sort of in more of a traditional software engineering role, but could you talk about sort of your transition into Cloud, from that background?
Kesha: Yeah, so I would say maybe seven or eight years ago, I was introduced to the Cloud. And like you said, I've been in IT for 26 years, and I really built the bulk of my career in the Java Software Engineering space. And I remember, and this will date me, I remember when Java came out, and I researched the language, and I felt like, "Oh my goodness, this language is going to take over the world. This is what I need to learn." And so I actually left the company where I was working, so that I could go to another company and learn Java on the job. And so the way I felt about Java back then, that's how I felt about the Cloud, maybe seven or eight years ago. I was like, "The Cloud is going to revolutionize how we deliver our software, how we support software, and this is something I need to learn." And so once I started with the Cloud, I've never looked back and every day is a new exciting day.
Adam: And you're an AWS ML Hero, so you're a machine learning Hero. Could you talk about sort of how you got into, I'm assuming you got into machine learning after you got into Cloud, or just tell me more about that, kind of how you got into machine learning?
Kesha: Sure. So maybe three years ago, I felt about machine learning, what I felt about Java and the Cloud, and I just realized that machine learning was just going to change lives in just ways that we couldn't even imagine. And so it was a technology that I wanted to learn. I always thought, being in IT, I knew about machine learning, I knew about AI but I always thought it was just this difficult technology that you had to be a research scientist, you have to have a PhD in order to even utilize it, understand it. And with the Cloud, I realized that wasn't the case. So I was able to really get started with machine learning because of the Cloud. And yeah, I just fell in love with machine learning and then I started sharing my lessons learned with others and that was a great way to become a part of the AWS Machine Learning Hero program.
Adam: Yeah. Could you speak to some of the projects, things that you've enjoyed working on?
Kesha: Sure. So my very first project, I think that project is my favorite project. So if you've seen the movie Minority Report, you remember this concept of pre-crime, where a criminal was arrested right before they committed the crime because they knew the crime was going to happen. And so, my very first project for machine learning... Okay, don't laugh. But all of my AI projects, I give names, okay. So I created Sam. And Sam stands for Suspicious Activity Monitor, but basically Sam was my version of pre-crime. And so I've searched and searched, with machine learning, you have to have good data in order to train these machine learning models. So I actually found crime data, stop and search data from the UK. And it basically listed out every time they stopped someone and whether or not that stop led to an arrest. And they tracked attributes of location, race, gender, age, and all of that information and so my very first machine learning project, was to build pre-crime.
Kesha: That was a fun one.
Adam: That's super cool. Sam is a little confusing with... We've got the [inaudible 00:04:32].
Kesha: I know, I was so upset when AWS came out with the Sam Service, I came up with the...
Adam: They always stumping on your project name. Yeah.
Kesha: I created Sam first, before that service came out. Let the record show.
Adam: Yeah. And did you partner with someone when you were building that out? Was that just sort of a pet project or...
Kesha: No, it was really a pet project. It was... Whenever I start to learn a new technology, I learn best by doing.
Kesha: And so I always try to pick a fun, real world scenario, that I can build.
Kesha: And so that was my first introduction. And way back then, there was actually a service called Amazon Machine Learning. Now that service is no longer available for new customers, but it was a great way for me to learn machine learning, because it abstracted a way, a lot of the deep technical details. And that was a great starting point. And so starting there, it just allowed me to start very high level, and then start to peel back the layers until I was ready to start writing my own custom training code using ML. So, yeah, it was awesome.
Adam: So do you feel like ML is as accessible as ever building out on AWS?
Kesha: Definitely. And I always tell people, I would have never really gotten started with machine learning if it weren't for AWS. And so another project that I built, you mentioned the Soda-theft Detection System, using AWS DeepLens and that is really the whole point of, I call it the DeepLens toy, but it's mainly for developers wanting to explore and learn machine learning. And so AWS, specifically, has really opened the door for just, I call it the everyday developer to pick up machine learning and actually use it in the real world.
Adam: Yeah. And can you tell us the project name for the Soda detection?
Kesha: Oh, that one didn't have a fun name.
Adam: Is it embarrassing? Okay.
Kesha: No, I didn't give that one a name.
Kesha: Yeah. I didn't give that one a name.
Adam: Yeah. That's understandable. So you have sort of this traditional software engineering background and you moved into the Cloud, and machine learning. Is there any sort of advice you would give someone today with a similar background? They've been a software engineer, they're a web developer, something like that. And they want to move into machine learning. Is there like first steps, things that you would recommend on that?
Kesha: Yeah, I would definitely say start by learning the machine learning life cycle. So as a software engineer, there's this proven life cycle to build software applications. So there's also a similar life cycle for building machine learning solutions. And so I think that's a great way to start. Just understand the life cycle and then become familiar with some of the terms around machine learning. Like I threw out there like machine learning model. I threw out there the word training. So just understanding the life cycle and the terms, and then just go build something, pick a fun use case and build it. Like, I always tell people there's nothing wrong with learning as you build. And really that's, for me, that's the best way to learn.
Adam: Oh, same. Yeah. I think most of everything I've learned in my career has been by building something to learn a new technology. And speaking... So one place people could to learn about machine learning on AWS or just AWS in general, A Cloud Guru, which you're an instructor. At A Cloud Guru. Could you speak, yeah... Could you speak sort of to your transition into teaching? How long have you been at A Cloud Guru and what did that look like?
Kesha: Sure. So I've been teaching much longer than I've been at A Cloud Guru. So let's say a little over two years, I made the transition to full-time teaching with A Cloud Guru. But before that, I would say 10 plus years I was teaching. And so, before I moved to A Cloud Guru, I was serving as a software engineering manager in the IT department at Chick-fil-A and I'd been there for 14 years. And what I realized early on there that as I began to move up the career ladder, my role became less and less technical. And at one point I wasn't coding at all. And I was really afraid that I was going to lose my technical skills. And so I thought to myself, "Okay, what can I do on the side, outside of my day to day job to keep my skills current."
Kesha: And that's when the light bulb went off that I can teach because as a teacher, one I already enjoy mentoring and helping people. But then as a teacher, it's going to force me to keep my skills current. And so I decided to start teaching part-time and I started teaching for the University of California. I taught in their Java certification program for over 10 years. Like Java one, Java two, object oriented analysis and design. And what I found that really, like I said, forced me to keep my skills current. And then from there, it led me to speaking at technical conferences.
Kesha: So whenever you attend a technical conference that speaker is like a teacher because they're sharing lessons learned about whatever technology they're using. And so I started speaking at conferences. And then from there, different organizations like Pluralsight, LinkedIn Learning, which was Lynda at the time, Manning, Udacity, they all reached out to me to start authoring these online courses. And then that just opened up this brand new world that I didn't even know was out there. And then from there, I think I did the Soda-theft Detection Project for A Cloud Guru, just I as a part-time consultant. And then I just, I really loved it. And so, a little over two years ago, I made the decision to start doing it full-time.
Adam: And did you sort of have dreams of being a teacher even before you got into tech? It was that always something you...
Kesha: I did. I remember when I was in high school, I told my dad that I wanted to be a teacher and his response to me was, "You'll never make any money being a teacher. You have to find something else to do with your life." And so, yeah, I found something else to do. I found another love, which was computer science and technology, but now I feel like I've come full circle and I'm a teacher.
Adam: And you're teaching. Yeah. But you mentioned mentorship and there's sort of some commonalities there in terms of teaching and mentoring, but you're an active mentor. I mean, that's very clear just looking through sort of, just going to your website, kesha.tech. I think it's very clear that that's a priority of yours. And I guess, could you speak to how you view mentorship in tech, both in terms of the importance in your career and maybe mentors that have guided you along the way?
Kesha: Right. So for me, mentorship is something that is very important, I believe, especially for women in tech. So I remember like early in my career wanting to have a mentor and just believing that I needed a mentor to help me progress to the next levels of my career. And I just didn't really see, and I didn't really have mentors and I didn't really see any mentors that looked like me. And so for me, mentorship is more about being that mentor that I never had. And then plus it just goes well with my personality because I enjoy helping others and sharing lessons learned. And I remember like midpoint in my career, I actually considered leaving tech because of some of the issues that I was going through. And I believe at that time, if I had a mentor that had already walked that path, it would've made things easier and better for me. So that's why I'm just really passionate about mentoring.
Adam: And I wanted to speak a little about that. It's 2021. I think technology companies are trying to make an effort to set up policies to improve diversity in tech, but that hasn't always been the case. And I guess from your perspective, you've been in tech as a black woman for 25 years, you've seen sort of any progress we've made. You've been on the front lines there and seen that. Could you speak a bit to sort of where we are in 2021? Are we making strides? Is it a more diverse and equitable industry today than maybe the beginning of your career?
Kesha: Definitely we're making strides. I'm really happy that there's a lot of focus and emphasis on diversity and inclusion and equity in tech. And it's a conversation that a lot of companies are starting to have and they're starting to pay attention. So I do feel, I feel hopeful about the industry and the future of the industry. I have three children, my youngest, as a girl, and she already calls herself a computer programmer because I'm her mom. And so, for me, I want her experiences when she is ready to start working in the corporate world of tech, I want her experiences to be different. So I am hopeful that she'll experience, her experiences will be better and different from mine, but I still feel like we have a long way to go until I can look at a company and I'm looking at the highest levels of leadership until I can see people that look like me, then, yeah. I still feel like we still have a long way to go.
Adam: Yeah. I feel like that's the disconnect. It's sort of a lot of publicly making, hiring practices more diverse. But, yeah, when you look at the people that are doing the hiring and the people making the decisions, or when you look in the sort of venture capital and startup world, how much of a monoculture and how everyone sort of looks the same in that space. I think there's so much that needs to be... It's easy from my perspective to feel like that things are improving, but I don't live your experience. And I've had a very different experience in tech obviously. So it's nice to hear those honest reactions of where are we today. Are we on the right track? Are we making strides? And I know you're one of those people that is pushing and making those strides and helping to improve that next generation of women in tech. Their experiences are going to be better because of people like you.
Adam: So you've also started something called SalaryOverflow, and I've seen a bit about that and I think it's sort of along these same lines. It's sort of trying to bring light to salary disparity. Could you speak about that a little bit?
Kesha: Right. So SalaryOverflow, that's another pet project of my... Have a lot of pets running around. But yeah, SalaryOverflow is an application that I created to bring transparency to tech salaries with the overall hope to close the gender pay gap. So in my career, I've experienced situations where I know my colleague who is doing the same work that I'm doing and who actually has less experience is making more money. And it's just not, to me, it's just not fair and it's not right. And so with SalaryOverflow, you can go into the system and see what others with the same title, same years of experience, the same location, what they're making. And so, my overall vision for SalaryOverflow, let's just say like, you have a job offer in hand and you want to know is this job offer fair? Then you can go to SalaryOverflow and find out if you need to, oh, well you always need to negotiate. I don't care.
Adam: Sure. Yes.
Kesha: You always need to negotiate, but now you know how much you need to negotiate for. And so that's the main reason why I create SalaryOverflow, but then there's also an additional reason. And that was for me to just play around with AWS amplify and learn some new technologies. So it was a win-win.
Adam: Yeah. I think I saw it's sort of a serverless stack. So that was sort of your first experience with amplify and learning some of that.
Adam: But yeah, it's nice to know the background. So I do the same thing. I mean, I think anytime I'm faced with some new technology that either a client wants me to build something with it, or I'm just interested in it, I sort of start up a new project and find something to build with it.
Kesha: Yeah. That's the best way.
Adam: So you've... Yeah, and you've had a few thousand submissions, I think, so far on SalaryOverflow.
Kesha: Yeah. Yeah. Every time I open the app and I see salary information, I just smile.
Adam: Yeah, that's awesome. You're doing your part. So you're also an Alexa Champion. Could you speak to, I guess, that program and what that means and just your experiences with Alexa?
Kesha: Sure. So I love Alexa and it's funny like around my house, I refer to Alexa, she or her, and my daughter's like, "Don't you mean it?" I'm like, "No, you could never call Alexa it, you can't do that." So Alexa was really, let me see, how many years ago was that? I can't remember how many years ago it was, but I just remember walking into my parents' house. And my mom had an Alexa, an echo device sitting on her counter and I thought to myself, "Okay, my parents don't even have caller ID I'm in tech. And my mom has this cool toy that I don't have." So that very same day I went home and I purchased an echo device. And then after that, it just, it peaked my curiosity to figure out how to build apps for Alexa.
Kesha: And what I found Alexa is really a great entry point to the Cloud because whenever you build an Alexa app, it's called a skill, you're actually using the entire AWS ecosystem to build that app. And so my very first Alexa skill used DynamoDB, Lambda, of course, Cloud Watch, IM. And so it was a great introduction actually to the Cloud and also to AI. And so once I built my first skill, I built several other skills, and then I think that just brought recognition from Amazon for the work that I was doing. And so they named me an Alexa Champion. And so I've spoken at several tech conferences about Alexa and the AI behind Alexa. And I just think it's just fun.
Adam: Yeah. And do you have like a favorite skill that you've written?
Kesha: Yeah. My favorite skill is called Word Jumble. And so basically it's where Alexa, like jumble scrambles, or jumbles a word. And then you have to guess what that word is. And so with that skill, I integrated so many different neat features and services. There's like a leaderboard using Amazon GameOn. I use poly, I use the notifications API. So yeah, that was a lot of fun building that.
Adam: Yeah. So you just, you named an AWS service I didn't know existed, GameOn, that's a new one for me. So I've got to look up GameOn.
Kesha: You can integrate leaderboards in apps.
Adam: Wow. So there's a service for that. That's news.
Kesha: There's a service for everything.
Adam: Yeah. There is. There really is. Love AWS. So I built some... Like I had a startup, still have a startup in sports technology. So, and really, we started like seven years ago before Alexa was even launched, building out tech to understand questions like about sports. So people asking how many touchdowns Peyton Manning have, that kind of thing. So we had a website with a search bar and you could type in and all that. And then really like Alexa came as we were building out this technology and it's so much easier to build out natural language understanding or whatever to build out that conversational, that voice experience. How do you see voice kind of in the future? I know there was back when I was much more involved with my startup, there was like this big kind of hype cycle moment where voice was the thing, and everybody was going to have an Alexa and everyone does have an Alexa, but I know it was really, really exciting. And then it kind of tapered down. How do you view kind of the future of voice?
Kesha: I think it's here to stay and it's just going to continue to integrate in every part of our lives. So for me, like my big thing now is I'm turning my home into a smart home. And so like Alexa controls almost everything in my home. I can vacuum, I can unlock the door. I can turn on the lights just through voice. So in the future, it's just going to be a natural part of our lives to control everything through voice.
Adam: You've got, you mentioned three kids, I think.
Kesha: Yes. Three.
Adam: So I guess my main question, I think if I got you on this podcast, ask you one question it's how do you do all the things that you do? I have few kids. I do nothing. I can't get anything done in a given week. And it's amazing to look at the things you're involved with. What's your secret?
Kesha: Well, there's never a dull moment. That is true.
Adam: Yeah, absolutely.
Kesha: I think my secret, I don't know if it's a secret. I've just, I've had to make sacrifices along the way, I come up with shortcuts and things like that. But to me it's just, it's natural now. Everything that I'm doing, everything that I'm working on, it just comes natural. And yeah, like I said, there's never a dull moment. The best times for me, I guess once I put the kids to sleep and the house is quiet, that's when I'm on my computer building, exploring and learning new things. But yeah, definitely I've had to make sacrifices over the years, for example, just sacrificing personal time and social time with friends and things like that. So there comes sacrifices in order to do everything that you want to do.
Adam: Yeah, absolutely. And what is... Is there something in the future you'd speak to? What are Kesha's plans? You're teaching at A Cloud Guru, anything big on the rise? Anything you have in the works?
Kesha: Well, just really more of the same and just trying to figure out ways that I can continue to mentor and help people along their journey in tech. For me, that I really believe that's why I'm here. And that's really what I enjoy doing. So anything in the future is going to still include mentorship and sharing lessons learned. So, I was thinking about this the other day. Okay. I've created Sam. I've created SalaryOverflow, like what's next? So first it was Java, then it was the Cloud. Now, it's machine learning. Okay. What's next after machine learning?
Adam: What is next?
Kesha: Yeah, I don't know.
Adam: Are you getting into web three? Are you going to be building stuff on the blockchain or what is Kesha doing next?
Kesha: That's a good question.
Adam: I feel like we could just watch your career and kind of figure out what's coming because you've been ahead of it.
Kesha: Yeah. I haven't quite figured it out yet, but that's funny that you asked that question because just the other day I was thinking, "Okay, like, what's next? What do you want to learn next?"
Kesha: You'll be the first person to find out.
Adam: Oh, thank you so much. And I might share it with everyone that listens to this podcast. Well, Kesha it's been so great. So finding you online, kesha.tech is your website.
Adam: You're on Twitter. Kesha Wills. I think all these links would be, we'll have them in the show notes and they're on your website.
Kesha: Okay. Awesome.
Adam: Thank you so much for coming on. It's been so great just to talk with you and just, again, yeah, this one's been on my calendar. I've been very excited just to meet you and learn more about your career. Thank you.
Kesha: My pleasure.
Adam: Yeah. Thank you to everyone that joined. We'll be back tomorrow and it's been a pleasure.