8 Bits

8 Bits with Cassie Breviu!

May 28, 2022 Pj Metz + Brandon Minnick
8 Bits
8 Bits with Cassie Breviu!
Show Notes Transcript

This week we are joined by Cassie Breviu!

Cassie is a Senior Technical Program Manager working on the AI Frameworks team at Microsoft. She focuses on the product ONNX Runtime, which is a product to deploy models cross-platform and improve inference performance.

Brandon Minnick:

Welcome back, everybody to another amazing episode of eight pits the show where we interview the people behind the tech to share their journey and hopes to inspire more people to do the same. I'm your host Brandon Minnick with me as always is my amazing co host PJ Metz. PJ, how's your week? I had COVID.

Pj Metz:

Let's go ahead and open with the truth. We are truth tellers. No. So I I had a weekend to myself. Last weekend, my wife went to visit some family and I was like, Oh, I'm gonna go there's this outdoor beer festival and I was excited about it. I was like, and it's outdoors. And then I was feeling really brave. So I went to an indoor concert the next day. And guess what, wearing a mask on the edge of the crowd. The whole time, a brief 15 minutes in the pit because they're they're a band that requires a mosh pit. The mask came off. I was really trusting my vaccines. And I trusted him a bit too much this bat is no joke. So caught it got sick had been isolating. I haven't really left this room that I'm in the office side of the room over there's my bed and it's kind of been just back and forth for the past six, seven days.

Brandon Minnick:

That's tough, but you look like you're doing better. It sounds like that feeling better.

Pj Metz:

drink Gatorade and lots of lots of chicken Lipton chicken noodle soup extra noodles.

Brandon Minnick:

That just sounds nice.

Pj Metz:

I don't listen.

Brandon Minnick:

I mean, I don't know about every day but I can get

Pj Metz:

adjusted doesn't want extra noodles. How are you doing? Oh, good have COVID Did you

Brandon Minnick:

know I was I was also outdoors. This past weekend. We went camping down in Sequoia National Park and Kings Canyon National Park, which are two adjacent national parks down in Southern California. So it was amazing. Really cool because sequoias got the big redwood trees. So yeah, it's like oh, gosh, no, you want a people holding hands around, it probably still wouldn't even make it around. Just oh my god, insane. And so. And then you juxtapose that with Kings Canyon, which is this huge canyon created by glaciers back in the day. So very much similar vibes to Xi'an if anybody has been designed on or even Yosemite kind of just you feel small. You're surrounded by these giant trees, these mountains coming up out in what seems like nowhere made out of granite. And it's just like, You know what? All that stuff. I thought matters. It doesn't matter.

Pj Metz:

It's so nice to have that place in the universe, like firmly secured like the first time you see the Milky Way and you're like, oh, like just nothing. Advice From that moment on, you're like, Yeah, I'm good. I don't I'm not worried about anything.

Brandon Minnick:

Yeah, so that was lovely. It was our first time taking a road trip in our in our Tesla, so figuring out how to charge it, when you're also staying at a campsite that's off the grid was interesting.

Pj Metz:

Fun.

Brandon Minnick:

Yeah, it basically meant we drove an hour back into town back into Fresno every morning, charged all the way up to 100%. That was enough to get us around all day. And then we do it all over the next day. So they should probably average channels on that thing. We were talking about that there. I've seen some mods out there that people have hacked together. But

Pj Metz:

just imagining like pulling a trailer that's got a whole like array of like, the trailer 36 feet long,

Brandon Minnick:

easy. Now the one I saw, it kind of pulls like almost like those barbecue tailgates that you can put that kind of pulled out. It looks like it reminded me of that, where you get to the campsite, then you pull this out of your trunk and kind of unfurl the array. But we made it, we survived. I mean, we you have to drive a lot around, especially these ones that are just so big. And so I don't even think that would have saved us because those, those guys used a lot of juice. But there's a lot of fun. And big announcement. We are recording this today on May 24, which is the first day of Microsoft build. It's the big yearly developer conference that Microsoft hosts, which really means it's the time when we at Microsoft debut a lot of our products. So the product that I've been working on for the last year of my life is the dotnet Malli community toolkit. It's finally live. It's been out in preview for months now. But we've just launched v1 This morning, really early this morning, because Gerald my buddy who releases it lives in the Netherlands. So as soon as he woke up, he shipped it. So beautiful announcement to wake up to. If you're a dotnet developer, check out dotnet Maui, which is also being released today. It's a way to build cross platform applications for mobile and desktop. And the best part is, it's all in C Sharp. And they're all using the native API's underneath. So totally native app that you can write in C Sharp running on Mac OS, Windows, iOS, Android, you can use the dynamic community toolkit that I made to make your life easier. Go check it out today. It's It's really amazing. It's been a lot of a lot of hard work from the team. So good to take a moment this this weekend to just breathe.

Pj Metz:

Finally, I just realized that I missed two days ago, we had a big thing get live too. We just had getting like 15 release on Sunday 22nd of every month we released a new version and we just hit 15. So we're at 15 Dotto and we're super stoked about it. We're gonna actually have like a several month long celebration. Cool stuff happening at GitLab as well. Man, we work for some good companies, huh? Also you have been working on that toolkit for such a long time. I'm so like, it's alive. You know what I mean?

Brandon Minnick:

Thanks. Yeah, it's like the first product I've birthed at Microsoft from literally nothing to starting conversations with lawyers on the legal team and engineers on the docs, team and mark

Pj Metz:

when legal talks. So I'm glad that you had a good one.

Brandon Minnick:

But anyways, we have somebody as a guest today who is way smarter than me, was on a product that's way cooler than what I do. And I'm so excited to have her on because one of the one of the first times we met she was actually keynoting at a conference. And I was like, Who is this person? So without further ado, Cassie, welcome to the show. Hey, thanks for having me. Thanks so much for joining us today. For the folks who haven't met you yet. Who are you? And what do you do?

Cassie Breviu:

Yeah. So what I do right now, I've had a lot of different roles with Philip probably talk about a little bit. But currently, I'm a senior technical program manager on a product called Onyx runtime, which is part of the AI frameworks org at Microsoft. And what Onyx runtime does, it's actually an open source product that allows you to do machine learning inferencing on multiple hardware platforms, like fusion providers languages, so usually, you know, you're building models in Python or R. And then necessarily, when you're deploying and going to use those models, you're not necessarily using them in Python applets, you might be using a C sharp, maybe you want to use Maui, like you're talking about. So this would actually allow you, our product allows you to do that. So it's a runtime that allows you to take these models that you've converted to an onyx format and make them run everywhere.

Brandon Minnick:

That's incredible. Is when it's so easy. be on the same team together at Microsoft. And like I mentioned one of the first conferences we ever attended together, you were literally keynoting. Yeah. But you're also still very new to your career, which was incredible, because the keynote was amazing. Huge, huge applause at the end of it. So let's, let's go way back. We love hearing everybody's origin stories on apex, because everybody has a different path into tech in into where their current role is today. So let's go way back. When did you first learn to code? And how did how did you get into all this?

Cassie Breviu:

Yeah, so I am one of those non traditional backgrounds. And it actually started when I was a data analyst at a company that was a third party administrator for benefits. And we needed to do this eligibility audit for insurance. And it was actually supposed to go to the engineering group, but we had some in house developers, but they were too busy. So it landed on my desk. And so it started with me having to do this large audit manually. And I didn't want to do that. So I started learning, Excel macros, or will Excel formulas, which then turned into Excel macros. And so I clicked record in Excel and recorded a macro and then viewed source and then hit play and saw that I had just created code because it's VBA. And I was like, What is this magic. So that was like, literally where my journey started was in Excel, trying to try to automate something I didn't really want to do.

Pj Metz:

I've met a lot of people who started out with like, I just need to automate some part of my job that involves data. And so then they start down that road, and then they're like, Wait, what's this, and then they like, they get like, distracted and start a whole new career.

Cassie Breviu:

And so then once I had like, automated half of my job, because I did so much in Excel, and so much with different formulas, I started getting macros that would call other macros and so like to run my, like half day work, I would just kick off my process, you know, debug it if it got into an issue. And then it was like, Well, I really liked this coding thing. And like, I think I just want to do this, like, I think I want this to be my job instead of using it to not do the stuff that I don't want to do in my job. And so that's kind of where it started. And then once I wanted to move, I knew that's kind of what I wanted. And I was like, Well, I kinda want to get into like a technical role. So I actually went found a helpdesk role at a small company that did non credit, continuing education. And from there, I worked in their helpdesk while I took classes at a community college. And so I actually did end up getting an Associate's in computer science degree. Now, this was before, like boot camps existed. And if boot camps around, I've kind of wonder if I would have ended up doing that. But I was looking for something that was not a four year degree, that was something I could do online. And, I mean, this was like 10, probably, like 10 years ago. And so that was like, now there's so many options, like, it's so much easier now. And so at the time, it was like, well, this seems like a good way to do it. So that's why I did it. And then, because I had put myself in a technical role I started getting exposed to, to everything. And because it's a small company, you know, you're you can be exposed to a lot more in small companies, because there's not like the sectioning off of roles, right, like, so I could see all of these different things. I was working with the developers, I was working with the business analyst. And then eventually, I moved into a business analyst role there. And I started designing product features for our product, and meeting with our stakeholders and customers and getting back those feedback and started creating these wireframes. And so that's what I did until I got my first development role. So I always tell people, well, one of the things that I think always worked for me was maybe taking something that wasn't necessarily what I thought was the perfect thing for me, but it moves me in the right direction. And then that helped me get to where I want it to be. Or I've been able to also like change my existing roles in to make them what I want them to be. So I've always kind of had, I guess, I've been called scrappy, want to call it to kind of make things what I want them to be. And then when I moved into my first development role, speaking of this scrappy nature of my background, I was actually a QA role. And I was like, Well, I don't really want to do QA I wasn't my goal. Their goal was to build applications, build software, I wanted to code. And they're like, well, so after my first interview with them, and I told them in my first interview, they changed my second interview and made it a developer interview. And so they actually interviewed me to become a developer. And granted, you know, I was brand new, so it's not like I was amazing, but I could do some things. And they ended up hiring me with the assumption that, you know, they could mentor me into becoming a developer. And that's exactly what happened. I worked in QA for a few months and then after that few months, I moved into a junior development role at that company.

Brandon Minnick:

Absolutely incredible. I love is one of the things I always say is a lazy engineers, a good engineer, because we look at these scenarios like, Oh, I got to spend half a day entering stuff into Excel. I don't want to do that. So let me let me just write some code, get rid of that. And so for anybody that's listening, that has a really mundane, boring tasks that you have to do, because we have, every job has that no matter how amazing your job is, I've learned, there's always going to be some mundane, say administrative stuff you got to do. See if you can automate it. Because next thing, you know, you might be senior technical program manager at Microsoft is right, I think what happened. That's the next

Pj Metz:

part, actually.

Cassie Breviu:

Eventually, I always say so the best way that I learned is like, I have a goal. And I just don't stop until I figure it out. So like, it was the same thing with Excel formulas and figuring out these macros, like I was looking everything up. And I was didn't really know what I was doing. I was copying pasting things in and trying to figure out what a class was, and all this different stuff. And so it's definitely that, that goal focused learning is is how I like to do it, because then eventually get something to work. And then I peel apart the layers and I figure out how it works. And and then when you're done, when you actually reach the goal that you've worked so hard for and probably wanted to give up many times, you realize that you've learned a new skill. And that's actually how I got into AI too, because now I focus on AI, which back then I was doing C sharp, full stack development and web development and stuff like that. And it was it was the same thing, which actually came from being lazy as well. Didn't want to assign helpdesk tickets anymore. So that was a machine learning model that I built.

Pj Metz:

And that's got, you've got a lot of fans of your desk as a hammertoe on Twitch said, I actually built my desk back in the UK based on her blog posts of her desk setup, using steel pipe to make a frame for monitors, lights, cameras, etc. So like, not only are you doing exactly what you want, and like making these great strides in your job and all this stuff, but you're inspiring other people to make better desks. So we applaud you. That's fantastic.

Cassie Breviu:

That's awesome. Yeah, that block has gotten a lot of that was actually from build, oh my gosh, how many years ago? I want to say two years, two to three years ago, I had posted my my setup. And like I got a ton of questions about it. So I wrote that blog really quick, because I had so many questions in the chat. I couldn't keep up with it. And so I was like, Well, I'm just gonna throw together a blog, and so people can see it. So I love to hear that somebody actually actually did it. Yeah, it was really nice. And the whole reason I built it was because I couldn't find a mount that supported all of the things that I wanted to mount to my desk. And so I took galvanized pipe and made my own frame so that I could sound like a lot of stuff. If you go on my Twitter and look, you'll see like I have so much tech on my desk that are telling

Pj Metz:

us before the show that you were like yeah, I just always want more and more and more and more. Oh yeah, for my computer,

Cassie Breviu:

the gadgets, I have all kinds of gadgets like I won't sit and list them off, but like it even things like I don't even need like I have this motion capture I'll show you one of these iPads like Leap Motion capture control. And essentially it has different cameras and sensors that allows you to do all kinds of cool things but one of them is to control things with your hands so you can feel like a magician so I didn't set it up yet but I'm like I really want to set it up because there's an app that you can add that will actually allow you to like control your monitor like with your hands like can do things like you probably wouldn't do every like Minority

Pj Metz:

Report like that's what I was thinking

Brandon Minnick:

like move windows around Yeah, yeah, zoom in.

Pj Metz:

I just I just want it so I can do MS Paint and like do it in the air and like it like the spray can and like I'm

Cassie Breviu:

actually I got it because I have this other gadget on my desk. Yeah, yes. Looking Glass. 3d. It's a holographic display

Pj Metz:

that looks like a 3d model. So like she held up like a picture frame for people listening to podcast and like there's a picture inside of it but it looks 3d Like it's got shadows behind it. But it's just a picture frame.

Brandon Minnick:

Yeah and as you rotate

Cassie Breviu:

like it things move in it as well. What Yeah,

Brandon Minnick:

wow, this is incredible. So yeah, again it's it almost looks like there's somebody inside of a shadow box moving. But what what is it is it flat? Like is there actually depth to that and real

Cassie Breviu:

depth so it is tricking you it's tricking you into seeing and hollow a hologram but it is a it's a holographic display and it looks like I know you can kind of see it here and they can't see it but it looks like the future to me like most what times when you think about I'm really into XR and VR and and all of that and AR and all that stuff too. And so really like most of the ways that we interact with virtual like 3d objects right now are really through wearing something Right, like you've put on. This is the first device that I've seen that I think does a really what good job and really like, I mean, it's still expensive, but more it's more accessible than I mean the same as like an Oculus quest, you know, that allows you to do 3d out that without putting anything on your face, like with being able to do that. And I've seen some really cool kind of like, projects with it that are actually like doing video calls, like where you can do video calls. In 3d. Yeah. So I really think this is the beginning of more of more 3d devices, at least I hope because. But yeah, there's, there's so many things you can do with it. So I got that in the Magic Leap, because I'm going to build um, you can also build things where you can control the 3d monitor with your hands. So like you'd have that 3d display. Leave, just like Ironman.

Pj Metz:

My whole life is lived through movie references, and we're getting closer and closer to the Riddler is invention from Batman Forever. Or Jim Carrey was the Riddler where you watch TV and it's like, 3d. Yeah, I'm just, it's all superhero movies in my head, I'm sorry.

Brandon Minnick:

deep cuts. It is.

Cassie Breviu:

And if you think about it, I think Minority Report was like the first time I remember seeing a truly like holographic display that you could move with your hands like, and I always think it's like is art imitating life or life imitating art because you see things in movies that like don't exist, but they're so imaginative and so cool. And because of movie magic, they can make them and I feel like, eventually some smart person like comes along and figures out how to like make it somewhat real. But it takes time for the technology to catch up. But I feel like sometimes just like the people that dream up in the movies are the things that then we try to attain it in tech, like the techniques, look at it and try to mimic it in ways that can be fun, real.

Pj Metz:

I fully believe in that pipeline. Like when you think about like Star Trek in the 60s and the gadgets that they had, what happens is people go, Well, what's keeping us from doing that? And what's cute, and so they try it and then they find great new ways to create tech for us. Oh,

Cassie Breviu:

my iPad, there's like, do you can find like, really? blogs that are like Star Trek invented all of this stuff?

Brandon Minnick:

I was just thinking that yeah, the cell phone. Yeah, we haven't gotten the teleporter yet. That's annoying. Yeah, come on. Thank you so much. I think this weekend.

Pj Metz:

Why can't I teleport to a national park? That's amazing. Oh,

Brandon Minnick:

goodness. Yeah. It's it's interesting, because I remember back in the day, there was a there's a while where 3d TVs were like the new hotness. And I remember that was one of the first because I was always a big technology guy growing up. I always buy latest gadgets and whatnot. Yeah, I just bought a TV before that wave. So I just started my first job ever out of college, use my signing bonus to buy a TV blew it all immediately, of course. And and then yeah, 3d TVs came out was like what? Like, I just spent all this money on a new TV. Like, just out of spite, I hope this doesn't become a thing, because then I gotta buy another new TV. And thankfully, it did well for me. So I didn't have to buy a new TV. But yeah, I I always said because I didn't enjoy wearing the glasses, like the 3d TVs you have to put on the special glasses sync up to the frame rate. And it makes sense. But I was like, Wouldn't it be cool if you could have a TV? That just was 3d. And that's what this reminded me of. That's what that picture frame reminded me of his. We're just looking at it right now. We'd have weird glasses on or anything special. And yeah, it's 3d. So

Cassie Breviu:

they do make a big one. It's just insanely expensive. I think it's like$3,000 or something like, but there is actually like, I think three different sizes. I have like the, like the smallest one. But yeah, they're relatively I think newer company. And they do have a large one. And I got to use it at the mixer or an MIT reality hat. They had a large one. And they had like a, like a spaceship thing. And you could pull it apart with your hands. Like without actually touching anything, and you could like move it around and play with it. So I always wondered why those the 3d TVs didn't take off. But if you think about there's so many things that have been created and failed, but then came back better, like even VR has failed multiple times.

Pj Metz:

We all remember that.

Brandon Minnick:

But this time

Cassie Breviu:

you know, it takes time for things to catch up for the technology, the hardware, really to catch up for what we can dream up.

Pj Metz:

The only thing I thought the 3d TVs were really really good at was you could set it up to where multiple people could be playing a video game on it at the same time. And they would just put all the images over each other and Everyone wear glasses and everyone got a full screen view. It was super cool. And it's literally a year after that everyone was just playing online instead. play multiplayer in person anymore. But you can't screen watch anymore because it's just your screen. And then finally I'd maybe be able to beat my freshman year roommate at Halo.

Brandon Minnick:

Right back, back in my day, we had to play Goldeneye one TV.

Pj Metz:

Yeah, it was a four by three 300.

Brandon Minnick:

And then you chop that up into four screens for four players. Yeah, you didn't get a lot of pixels back in the day had

Pj Metz:

the tape cardboard on the line. So.

Brandon Minnick:

Oh, goodness, so. So Cassie, here we are in your journey. So we are now working in QA working towards this developer job. What happens next?

Cassie Breviu:

Yeah, so once I finally got the role, like the engineering role, I was so excited. But I was also like crazy imposter syndrome. And I didn't even know that's what it was called back then. But I definitely had that I was on an all male dev team, that I was actually it was a law firm that I worked at an intellectual property law firm. And many of them had been in tech for a very long time, most of them were senior level. So I was super intimidated as well. Like, I just remember my first stand up, and I probably like a deer in the headlights. So I was like, Oh my God, how did I get here? Well, and then, and then eventually, after a while, you know, at all, it was fun. And they I had some great mentor there and all of that. But so I started learning, I started building different applications. And the thing that was really cool about there is we built stuff from like, I would get in like a business, I was building apps for like internal lines of business. And so they would need something to solve a problem that they had to do manually. And so then I would have to go and kind of like I would design it. And then I would build up like the database, I build out all the. So I did like all SQL stuff, which was and then I did, you know, the C sharp back end. And then I did the front end with JavaScript. And like, I was still doing the whole thing. And then like demoing it to the the users in the company to see if it would work and what they need to change and all of that. So the thing that was really, really cool about that is a lot of times when you go into tech, and particularly now there's really a lot more like specialized roles. And there I was getting this like broad scope of everything. So I was very much immersed and was able to learn so much. We're all hosted on prem then as well, because cloud wasn't what it is today. Now, either. And also law firms had been slow, in general to pick up cloud, but so I just kind of kept going. And I kept building things and, you know, breaking things on the way as well. Because of course, there's always like the legacy applications that you have to kind of like reverse engineer and figure out. So we got to work with a lot of different languages because of that, as well. So I still think I got like one of the best first roles, at least at least for some free people that learn like me who like learned by doing, because I just got thrown into like all of this information. And I got to learn so much about the full process. And and yeah, so that was really cool. And then one of the things so I was in that and I wanted to learn artificial intelligence, like it was something I had, like seen, right, like, and I was like, What is this cool thing? Like? And I just thought it seemed I didn't know where to start. And I thought it seemed Oh, it's must be too complicated. And what happened is I went to a conference talk at that conference, actually the one that you saw me keynote at. And this was before I was keynoting. This was years before that. And I went to a talk and they went did a talk on how to survive the Titanic, or would you survive the Titanic? Sorry, not how would you survive? It's in a dataset. And they went through the full machine learning model model building process through this thing that's now called designer, Azure Machine Learning Designer. And I was like, I get it. And like, I mean, not like I understand all the machine learning and not what wasn't like Neo and like The Matrix. Forward is really what happens, and an understanding of what the data science process was. And so as I said, we're talking about earlier how lazy we had to assign these helpdesk tickets. And what would happen is, so we had a help desk there, but they didn't know because we had so many different in house built products. The helpdesk didn't really know what would go where and they couldn't necessarily tell from the ticket. So it would go into this like bin of A, that would then we'd have to sift through and figure out and nobody wanted to do it. So the developers would have to spend a week like maybe I'm every month where they would just have to have that distraction like if something came up because because with a tickets like this, they have to be triaged quickly, you count them sit, so you'd have to break your workflow and as you know with you In engineering and with building things like you can get very into things. And it's hard to once you break that concentration, sometimes really hard to get back in. So it was a very kind of, like, frustrating thing to have to do. And so I was like, Well, I think I could use machine learning to take the text of this ticket and figure out who it should be assigned to. So I don't have to do it anymore. And so that's actually the first model that I built, was how was assigning helped us tickets. And, and the way that I did it is we had our helpdesk ticket was ticket system, we hosted the data in house, but the software was purchased, I couldn't actually integrate into the code, I could only use the database in order to create events to do machine learning. So I, I had built this in the cloud now. So I've used this drag and drop. So if you're not familiar with Azure Machine Learning Designer, it's like a drag and drop visual tool that allows you to like build machine learning models, without writing really much code, you can write code, but you don't necessarily have to. And so I had hooked up to the database and had a trigger that would kick off the workflow if there was a ticket that was assigned to this bucket that we had to triage. And then I would take the text of the ticket, and it would influence on who it should be assigned to it would take the top three, and then it would email them post it to a team's channel. And it would say sincerely, the machine overload Lord. I didn't really tell anyone I was building this, like I had, I've kind of like started, I was like, hey, hey, I need access to this database. It's like why I was like, I just wanted to do something like I wasn't really telling people what I was doing, because for some reason, I didn't, you know, I didn't know if I was gonna be able to do it. And I didn't want people to get like excited because if they don't, what if it doesn't, right, all that stuff. But people started getting help just to get signed signing. The first model that I built,

Brandon Minnick:

like all of a sudden Cassius productivity just went through the roof, and you're deciding tickets within seconds of them being opened.

Cassie Breviu:

Yeah, it's great. Like AI was no different. For me, it was just the same way that I had done everything else. I had got a goal figured out something I wanted to build, broke it down into smaller pieces. researched a lot. But then, so then what happened? Because now you realize, like, I built a model, but I haven't used Python yet. Right? Like I haven't actually in at this point. I haven't even written any Python code yet. Like in my development career, like I had not done Python. It was all C sharp. I still love C sharp. But I guess TypeScript makes JavaScript bearable. So I'll say well,

Pj Metz:

come here for your hot languages, y'all.

Cassie Breviu:

Where was I? Oh, right. So then they're like, Well, this is really cool. But we want we want you to do other stuff with machine learning. Like they were actually really excited and really supportive, which I think was one of the great things about that place is they were always like, open to new ideas, and trying new things, which isn't always the case at companies. And so that was one of the things there that was really, really cool is like, they would be like, You know what, I think that this is useful. And I could see value, like, why don't you take some time and figure something out. So they wanted me to be able to build it on prem? Well, this tool was on the cloud. So I was like, I only know the models with this cloud tool. And they're like, well take a couple weeks and see if you can figure it out. And I was like, Okay,

Pj Metz:

just a few weeks. That's all you need.

Cassie Breviu:

Yeah, I know, it's like, you can get something to work in two weeks. I didn't understand all of machine learning in two weeks. But so I reverse engineered what I built in the cloud, doing watching some different stuff on Coursera, I watched the supplied AI course on Coursera. And I kind of just took different pieces of different things to figure out what I needed. And then put it all together and then looked at how I had built it in the tool and reverse engineer that and I ended up using scikit learn so I wasn't doing deep learning and I was doing classical machine learning. And yeah, I ended up re building it. But I mean, I put in a lot of hours, those two weeks because I was super determined. And I was like, I want this to work. So I was like really excited about it as well. You know, sometimes when you're like building things, there's like a push or a pull, right? Like you have to like push yourself to do things you don't want to do. But when you really want to do something you like it pulls you to the point where you can't stop until you done at least that's how for me so that was like one of those things I was putting in like a lot of hours just because I was obsessed with getting it to work. And I did eventually I did get it to work. So that's

Pj Metz:

at this point, Cassie that like if you get interested in anything given enough time, you will you will be able to do whatever it is you want. You are intimidatingly smart, and I am like terrified that like I'm on this like it's a virtual stage but like I'm scared that just you're really cool. Like this is amazing. Don't mind me. I'm fangirling here. I'm very excited to have every story here like so like I just started researching and like when I start researching, I get distracted and then suddenly I'm watching like the new Thor trailer over and over again instead of actually. So like, you're just what I think you're showing is like, if you really stick with that, first off, that push and pull example, is perfect. Because there are times you have to push. And you have to say, I just need to figure out the next thing, and I just gotta get through this. And there times you can't think about anything else. And using both of those ways to your advantage is really important.

Cassie Breviu:

Yes, totally. And some like this was really hard to like, I'm not I don't, I don't think I'm extremely smart. I think I'm just extremely determined. And I don't give up no matter like how many times and I know that sounds like really cheesy, but that honestly is like, I will literally code till my eyes hurt and like it hurts to blink because I've been staring at my screen for so long. Like when I do something like this. And I actually went out with a friend like that Friday, like, I think it was the first week when I was like learning everything. And literally, I was so tired, that I like could hardly form sentences. And I just told her, I was like, You know what, I'm sorry. Like, it's gonna be hard for me to find the right words, because like, my brain is so tired right now. Because I've been so like, focused on this thing. But like to talk about your distractions comment, like, I get like that, too. I just depends, like, I have to really, really like want something and then I don't get distracted. But like, I have a harder time when it's something that I have to like make myself to then it's a lot harder to do. So like it was really hard. Like, you can look back at it now and be like, Oh, I did this thing. And it's like, but I worked insanely hard. And it was like, really peered determination. That's really all like, if you're determined it you can. And you will.

Brandon Minnick:

Yeah, that's that's one thing I love about software, was actually chatting with a guy at the campsite this weekend, who used to be a software engineer, and now he runs a campsite. Which is not uncommon for a lot of folks in tech to eventually retire from tech, like, retire retire from tech. And yeah, he gave me a hard time. He's like, Oh, how long have you been doing this? Like 1015 years? Like, okay, so you're not there yet? Yeah, because I was like, I love to write code. And one of the reasons I love it is building on what Cassie said is you you can do anything like there are limitations. You know, there's, there's physics involved eventually. And you can't do anything anything like that's probably why we don't have teleportation yet. But you can essentially build anything you want. Writing code. And there's, I've, I found a similar thread in my career to Cassie, that, there's always a way to do something and rarely can just not be done. And so yeah, if you just keep trying stay focus that you'll eventually figure it out. It might not be easy. But that is one of the favorite things that I found when writing code is eventually I'll be able to fix this bug, eventually I'll be able to implement this feature because there's, there's got to be a way to do it. I just need to figure it out. And that's also I found leaked into the real world, like something in the house breaks, like I had to fix the hot water heater a couple of weeks ago, because all sudden, we ever taken cold showers. And that's not fun. And that's because this thing is designed to work in a specific way. If I can just figure that out, then I can figure out what's broken. And then I can hopefully by that replacement part. And so I love that advice. Just in general, there's there's a way to do it. If you just stick with it. And kind of Yeah, break it into smaller parts. towards that goal. Yeah.

Pj Metz:

Yeah, it's not it's not fixed the water heater, it's okay, find where it's stopping, and then look online, similar and then find that replacement part like an ace or Lowe's or whatever, and then figure out how to like there's steps in there and knowing what those smaller steps are, makes the big task manageable. So yeah, Cassie didn't go, well. I'm gonna make a I was like, oh, I want to automate a thing. Okay, how do I do that? Well, I saw this, this presentation at a conference. And I took that home and I started working with it, and I started finding things. So it's never, you don't accomplish the big thing without the 1000 steps leading up to it. And speaking of 1000 steps, here's the first step towards us getting sponsored on this show. If you're hearing my voice, that means you've been listening to or watching eight bits with Brandon MPJ. And we're here to talk to you about your product and how it can help you in your life by to do whatever your product does. So if you're an avid listener of the show, or you watch us on Twitch Then you will know that your product, your product is right for you.

Brandon Minnick:

That's right. We are just a humble podcast that still has to pay the bills. So if you have a product that you'd like to, for us to feature here on eight bits, send us an email Hello at eight bits.tv. And we'll be in touch. Now, Cassie, where we left off. You're, you're you're changing the world, you're doing anything you want, you're essentially seeing the matrix and making it bend to your will. But selfishly, I know, I've known you for four or five years now, specifically, because we've worked together Microsoft, and we haven't even gotten to that part in your story yet. So where do we go from here? What's next?

Cassie Breviu:

You mean? Like, how did I get from there to here? Or where I'm going now? Because I never have any idea. By the way.

Pj Metz:

How did you get to Microsoft? Your memory and

Brandon Minnick:

washing, I woke up?

Cassie Breviu:

You know, I was like, I like AI. And they're like, We like aI too. And I was like, Cool. Let's do it together, then. Yeah, no. No. So that's, that is one thing that I think is funny. Looking back now, like, I can say things very, like, it seems so streamlined. And it makes so much sense. And, and but it's the hindsight 2020 thing for sure. Because like, I did not know, I just knew that I wanted to do like, I just like when it started with the development thing. I'm like, I want to do something with development. So I got a job and, you know, help desk and then it's like, well, I know I wanted to do this. But I actually really liked the business, the design side of it, like when I was doing that product, feature upgrades and stuff like that, like I really liked this. And I almost changed into doing more of just that instead of moving into an engineering role. But I was like, No, I'm going to stick with my original plan or what I really wanted. But yeah, I really feel like it's one of those there's this like Fiona Apple song about the roll just rolls out in front of me. And that's kind of how I feel about it. It's, it's like, I know, I have ideas of things that I want. And it's not that I'm not goal setting, because obviously I set goals, and I've created things that were allowed me to get new skills that allowed me to progress, my career. But I didn't start out, you know, 10 years ago, or whenever it was, like, I'm gonna be, I'm gonna work at Microsoft, I never, ever in a million years thought I would have ended up here, I feel so lucky to have ended up here. But it was just and so honestly to the point because I don't have a traditional background. Right. So talking about how I got here, I never actually really applied to big companies, because it was a lot of times in the HR process, when you don't have a bachelor's degree, like you're just they just you never get called back. And, and even like with the small companies that I worked at, a lot of times I go through recruiters, because recruiters go straight to the hiring manager, and it skips HR. And so this gets HR, they're not hiring managers in tech know that it doesn't matter if you have a degree or not. And so that was like one of the ways that I had to kind of like work around the fact that I didn't have a traditional education. And then I would do temp, I would do temp to hire because then they can try it. Try you out, essentially and see if you work with it. So almost everything I did before, Microsoft was tempt to hire through a contracting agency, because I had to go around HR. So I never actually would have applied to Microsoft to be honest, but not because I wouldn't want to I just would assumed that they would never have talked to me. And then I was at that conference that I ended up talking to some Microsoft people the year after I had wanted that talk and I was having issues with hosting stuff in Azure, with just specifically what I was doing. I was doing okay, so now the funny thing is now is like, I looked back and I was doing ml ops, but no lapse didn't really exist yet. I was working with operationalizing models in ways for a startup application that I had been working on. And I was running into a lot of issues. And now looking back, I'm like, well, that's why I was having issues. It's because there wasn't a lot of people doing that quite yet. At least not at like in small companies. Right. So anyways, they asked if I would be interested in interviewing, and I was like, Yes.

Pj Metz:

Really? Talking to me, or is there another Cassie? Yeah.

Cassie Breviu:

Yeah, so that's that's how I ended up here. And, and I'm eternally grateful for the path that I've had. And I think a lot of it just came down to putting in the work. And but I know some people are much more like planned and they're like, they have all these goals and like vision boards and stuff like that. And I think that's amazing, too. And I think you get amazing results that way. I think everybody's just kind of different and there's not like one right way. There's not one right path. It's everybody kind of has to find their own way.

Pj Metz:

I guess. We have three totally different paths on the show. Brandon who always talks about how he went to traditional route But like he like your degree, Brandon was computer engineering, computer science hardware, computer hardware.

Brandon Minnick:

So designing microprocessors in college. Exactly. So

Pj Metz:

yeah, so not a sophomore. So even your traditional path isn't as traditional as some people who literally major in being a developer, you know what I mean? So like, every path is going to be different. And Cassie, I love that you said, like, everyone's got a different way to get there, it's just about what works best for you another person's way might not be the way you do it. And so we can't sort of, I think the idea of exposing people to multiple paths is really important. And that's why I love hosting the show. Because I mean, I get to meet so many people, and every single person's path is totally different. You got to trust it, and you got to put in the work, it's not just gonna happen to you, you got to put in the work. And, you know, having a degree is not the only way to do it. So don't feel like you have to get a college degree to make this happen.

Cassie Breviu:

Right, you do. But you do have to work hard either way. And I feel like one of the things I hear so much from people, and you mentioned something about it, like, you know, it's not easy that everyone so many people, they're like, I want to learn to code, like, how did you do what you did did it and I'll tell them, like how I did it. And there's, I will say a large percentage of those people don't necessarily follow through. And that's one of the big things is it's you have to understand like it is not going to be easy, and it's not going to be straightforward, it is going to be really hard. And that's what makes it great. Because also like if you think about like development, like PJ, now that you've been in a development role for a little while, like you know that you get really stuck on things, even after even if you've been doing development, you still get stuck on things, and you still have to put that hard work in. Stopping hard, that's the other thing that I feel like people should realize is like, if you don't have the hard work to put in to get there or to like learn the skill, then you're and if you don't like that, then you might not like the coding side either. Because it's very much like, yeah, you have to have, you have to work hard to do it.

Pj Metz:

And not only that, you don't have to be able to code in order to work in tech. There's a lot of great opportunities in tech that don't require coding, even my job. It's not I don't code day to day, like I'm making tutorials for like students, and I'm working with education and like, my I don't write stuff for GitLab. But I need to know a little bit for my job. But I know people that work in GitLab, that don't code at all. And it's not just people who work in like legal or finance or department require code. There are opportunities. And there are things out there that you have no idea what the job title is. But it might be an opportunity for you. If you had been like PJ, you're going to be an education evangelist, I would have been like, that's made up. That's not a career there. And there are other people and like you couldn't convince me otherwise. But like, here I am. It's, that's that's one of the things that I try and impress on my students that I talked to all the time, it's like, don't worry about being a developer. Like if you want to work in tech, get some skills keep working. And it's like what Brandon said earlier? And what you said earlier as well, Cassie, it's like when you come upon a problem, work to solve it. And it's sometimes you go around it, sometimes you go over it, sometimes you go under the problem. But solving the problem, there's a million routes to do.

Cassie Breviu:

You know, and I would like to say to now I see a lot more job postings that don't say degree required. And so if I look back, I think that I would have been less self conscious about the fact that I didn't have the degree and trust in the fact that I had the right experience. And so like, that's one of those things where like, I was probably cutting myself short because I had like this insecurity or like imposter syndrome about having the right background. And so I think that's one thing too, that I feel like people should realize is that like you should apply for the job that you want, even if you don't think you have the right background, or even if you don't have a degree, and also look for those job postings. Like a lot of them aren't saying they require degrees anymore, because they're realizing people can can learn these things other ways. So I just want to make sure that what I said didn't discourage anyone from like, applying for particular roles thinking, Oh, I have to get recruited. It's like no, I just that was the thing that showed me that I apparently didn't necessarily matter that I I didn't have a degree for that role. So I think it's an important thing to think about and to understand.

Brandon Minnick:

Yeah, let's seeing like you said, Cassie, the, especially the big tech companies like Microsoft, Google, they're all dropping that bachelor's degree requirement now, which is so good, so freeing will bring in, it's just gonna get more talent and more people from different backgrounds, which makes everything so much better. Yeah, but, Cassie, we only have 10 minutes left. I don't know how I don't know where the other 50 minutes went. But you are doing some amazing things right now that I want to make sure we show off. So where should we start? Okay, you've got a new product coming out.

Cassie Breviu:

Um, I've been working on I'm like, which one do I start with? Huge, obviously on it. So Onyx runtime is a product that we work on. And I do all of the Onyx runtime YouTube channel of videos and content there. And so there's a lot of different things there. And I post about every two weeks at least I try to post every two weeks. And I have one coming up like in this week that I am hoping to get ready and post around something really cool that Onyx runtime data. So we've partnered with Epic Games, to create a neural network inference plugin in Unreal Engine five. And so what this plugin actually does is it allows you to run neural networks or deep learning models within your unreal games. And so there's like a lot of different ways where that's starting to change everything, it's still kind of new, a newer area in the way where you'll see like the big companies like I don't know if you saw the one that NVIDIA did with GTA for photorealism. Or did you see that one?

Pj Metz:

That's just I'm just thinking about all the Unreal Engine five stuff that I've seen lately. And I'm convinced the uncanny valley is about to get shattered, like I can't tell, but I saw one it was like of a Japanese train station. And I felt like I was in Japan. I was, there was amazing. Absolutely,

Brandon Minnick:

let's just call it real engine. Now. They've just really made they made reality. It's incredible.

Cassie Breviu:

They did. Oh, surprising. When you said something about how like, there's still limitation with programming because of physics. Well, in virtual reality, you get to control the physics. And so anything is possible in virtual, right, you literally can create anything that your brain can think of, which is why I love XR and VR so much. And one of the cool things about applying AI into gaming, is that it opens up kind of this new door, and it's very experimental right now. But I feel like we're gonna see a lot of really cool leaps and bounds in that area. And so that's what I'm really excited about for the the new plugin because it makes it so you can use Onyx runtime as your inference engine. And then it has like open CV is a helper, which is another experimental plugin that allows you to do different computer vision tasks within it without having to create the plugins and integrate the open source libraries yourself into the Unreal Engine. So that's something that I'm working on, I'm really excited about and the the thing within it. So like, as I was saying, you know, like there's like the photo realism, one of the ones that I did that you'll be seeing soon is the style transfer. So are you familiar with style transfer, like aI models? So I bet I bet you've seen it, where you had a picture and you had like a Van Gogh or a Monet or something, and it made that picture look like a Van Gogh or Monet, right.

Pj Metz:

Okay. Yeah. Yeah. See, notice? Yeah,

Cassie Breviu:

that's essentially an AI model that's trained in a particular style, it can take your image in, augment it, using something called the Gann. Again, network, generative adversarial network, and then create a new image. Now, though, does the thing that I built does this an unreal using Onyx runtime, but it applies it real time to your game. So you can make your games like in the style of whatever the model has been trained. And so Onyx runtime has this open source library or open model Zoo. So there's tons of different models there that you can just go grab and use. And we have like, like seven different style transfer models, you can just go graph. And so I grabbed those, and I started playing with them and unreal. And this, this took more than two weeks to get to work, by the way. But it did work. And it's really cool, because you can look make it look like a painting, you can make it look like a mosaic. But it's still the vote, the quality of this experiment that I did is still not quite good enough to put in like a production. But I'm excited to release the code and like see someone like some, someone that's smarter than me go figure out how to make it higher resolution and make it more realistic. But it is something that is using Onyx runtime and inferencing in the game at every single tick, which is like kind of insane to think about like that would not be possible with older hardware at

Brandon Minnick:

all. Wow. So So if I understand this correctly, I can almost reskin a game using this. So definitely developers let me add a mod. I can make it look like I'm anywhere real. Right? Super

Pj Metz:

Mario World is now inside of Starry Night.

Cassie Breviu:

Right, exactly. Here. I'm going to show you a quick little sneak peek here. Oh my gosh, yes. With an example. So I have this really basic, unreal scene. So you can see it's just like grass and like cement walls, some stone and some bricks. And then when I would hit play on the Unreal scene, this is applying a picture called rain princess. So if I go to the Onyx model, zoo, I can show you what the actual model was trained on. And it's really cool to see how that happens. But this is like what it was what it would look like and then when you're able to walk through the world and mighty This is a very, very simple game because I just use like the base one from Unreal. And then went in and tried to figure out how to how to make that work.

Brandon Minnick:

Incredible. Just to describe it real quick for folks listening on the audio podcast, what we're looking at is Cassius essentially built a 3d world, there's like a brick or concrete fence, there's almost looks like a, an opening to a well, and some boxes around this will say a backyard that's fenced in, and then she's applied this texture to it. And now all those things are still there, the walls still there, the will still there, the boxes are still there. But the colors, the design is totally different. And Cass is pulling up the picture that it's based on. And it is on, as I say, unreal, uncanny how similar those textures became to how this model has been trained on.

Cassie Breviu:

So like this is this is that image, right. And this is why it's called Green Princess model of this image, that is also like it's what the this, this painting was called. So when you look at that, and then you look back at what this created.

Pj Metz:

Yep, impressionist painting of a woman standing in the rain, and applied that style to the objects inside of this, this is amazing.

Brandon Minnick:

Yet everything like the grass has,

Pj Metz:

I love seeing this bleeding edge, see.

Cassie Breviu:

So if I go back to the mosaic example. So this is the mosaic painting. And if I go up here, you can see how that that has translated in the example of how these models were built, which are built with pytorch. And then if you go back to the example you can see, so like this is an example where like, you can see everything, but the resolution wouldn't be good enough to like the rain. For instance, when I feel like it's like almost good enough where you could actually probably like, push that. But the mosaic one, it's like it's there, it's really cool. But it needs a resolution update. So I've been trying to think about some different ways that I could improve the quality. And I think I have to go back to the model itself and think about training it. One of the things is it's most computer vision models that are open source slot, and they go to 24 by 224 for dimensions. And this unreal output is a longer triangle or a longer rectangle. I mean, and so one of the things I think is that that translation might be causing it. And so if I were to try to train a model with a higher resolution, I'm wondering if it would improve that, but But either way, it's just really cool to think about like, this is just that first example of something that you could do and see where it's gonna go.

Pj Metz:

Oh, man, I can't wait.

Brandon Minnick:

It truly looks incredible. The other scenarios running through my head right now just sound amazing where anybody can just say, hey, here's some artwork, model a game and all its textures off of this insane cast. We only have about a minute left. Thank you so much for coming on. Before we leave. let folks know. Where can they find you folks want to learn more about Onyx? For folks who want to follow you on Twitter? Where can they find you online?

Cassie Breviu:

Yeah, so my Twitter handle is @CassieBreviu. And that heart probably hard to spell but not going to try it, you'll have to go look it up. And then the or I'm not going to spell it out. But then the other way to find me is the Onyx runtime Twitter account, I run that one as well with some of my teammates. And then also the YouTube Onyx runtime table or channel. So it's just YouTube slash Onyx runtime, oh 1x ONN X runtime. And that's where I post lot of my content. And also I pay attention to what people are asking for too. So if there's like something you've been trying to figure out and you want a video on it, and you ask for it, I'm not going to promise but it will be definitely something that I'll look at as an indicator of things that people need. So if you go to the Onyx runtime YouTube channel now, you'll you'll see videos on just different ways to actually build different types of solutions, including some onyx runtime with Xamarin, which I know Brandon loves to spend time for mobile. And specifically, we have a Xamarin package do that and C sharp and JavaScript so it kind of shows you how you can do inferencing and all these different languages and then the Unreal Engine One is here, but the actual demo isn't out yet. It'll be out tomorrow hopefully

Brandon Minnick:

amazing. Cassie I'm so glad you were able to make time for us. Your your story is truly truly inspirational. It's amazing to see that you can start by just and centrally being lazy and not wanting to read spreadsheets anymore to now creating machine learning models that game developers can use to transform the industry. Thank you so much for your time today. And thank you for joining us for another episode of eight bits, and we'll see you next time.