Kristina Hoeppner: 0:04

Welcome to'Create. Share. Engage.' This is the podcast about portfolios for learning and more for educators, learning designers, and managers keen on integrating portfolios with higher education and professional development practices. 'Create. Share. Engage.' is brought to you by the Mahara team at Catalyst IT. My name is Kristina Hoeppner. Today I'm talking with Mario A. Peraza from Stanford University, who is completing his Masters of Science in Mechanical Engineering with a specialisation in Design Methodology. And currently still a student but only for a couple more days when the podcast airs because he is preparing for the commencement ceremony and graduation now. Congratulations on having completed your degree, Mario.

Mario A. Peraza: 0:55

Thank you. Thank you.

Kristina Hoeppner: 0:56

Thank you so much for the chat today, Mario. Helen Chen, also from Stanford University, and she works in the Integrative Learning Portfolio Lab with you, she pointed me to one of your LinkedIn posts on artificial intelligence and how you use it in portfolios. That, of course, is a hot topic. And so I wanted to have a chat with you.

Mario A. Peraza: 1:20

Yeah, I'm super excited to talk about it.

Kristina Hoeppner: 1:22

Before we jump into the AI specific questions that we wanted to explore, can you tell us a little bit about yourself? What do you do at Stanford, and what did you do before?

Mario A. Peraza: 1:34

So here at Stanford, I'm a Master's student in the Mechanical Engineering Department only for a couple more days, like you mentioned[laughs]. What I've been doing is really focusing on design and engineering work. I am super big into human-centred design, working with different groups, whether it's projects or processes are trying to figure out how you can really integrate design and engineering together in a cohesive way. But as you said, I'm also part of the Integrative Learning Portfolio Lab and through that I get to run workshops and be more of an educator, which I super enjoy doing. And I've learned so much being a part of that.

Kristina Hoeppner: 2:09

Thank you so much, Mario, when did you actually create your first portfolio yourself?

Mario A. Peraza: 2:14

That's a good question. I believe it was around 2020, during COVID times because I had a lot of time on my hands. This was about sophomore year for me in undergrad. It's really the time when I was like, 'Okay, I should start getting a job.' So I created different versions of my résumé. And really, what I saw was, I was writing words about all the projects I've done. I'm a very project based person, just because I'm a mechanical engineer, and I have a lot of, you know, pictures and CAD files, 3D models, I want people to see. I can't do this with a résumé. So I created a portfolio on a free platform, and that's kind of how that started.

Kristina Hoeppner: 2:51

Did you learn about portfolios at your university at that time or did that really just come out of interest for you?

Mario A. Peraza: 2:59

I think it just came out of the need to have somewhere to show all of my projects. I knew about portfolios, probably just because other people in my class had them. I had a lot of senior friends at the time. And some of them showed me their portfolios, and I was like, 'This is a cool platform, I could just use this.' So I just started experimenting with the different website making tools. And that's kind of how that came to be.

Kristina Hoeppner: 3:24

Now you're not just a student at Stanford, but also student ambassador and intern at the Integrative Learning Portfolio Lab (ILPL). What does your role entail there? And how did you actually even get involved with that group?

Mario A. Peraza: 3:39

How I got involved with - I think I just saw the posting online. I was looking for something to join, something to be part of at Stanford, and I saw the posting and I was like, 'Oh, I have a portfolio, maybe I could do this?' I mentioned running workshops, which I love. I try to be an educator and try to improve my teaching skills as well. So I thought that would be cool to get involved with. But what my role really entails is kind of just like a lot of different things, managing the ILPL website, creating the Lab's YouTube video series, posting on LinkedIn. So if you go to there, those are pretty much all my posts that I've written and helping run the workshops, at least two a quarter is what we've aimed for. I would say that the most important role is the workshops because we typically try to find underserved communities on campus that I myself am a part of, and I can partner with them to tailor these digital presence curation workshops to them. So I've run a couple of queer community workshops, talking about what it means to be queer in a professional environment, first gen(eration) and low income centre on campus, we work with them, and you know, digital promotion, online presence might be a very new thing to them as well. So we try to tailor the workshop around that as well. But then also just for general populations. We run grad community wide workshops. Yeah, I think that's just been the super fun part of being a student ambassador.

Kristina Hoeppner: 4:56

Yeah, quite different also from your regular coursework, so it's really good balance to gain some additional skills there.

Mario A. Peraza: 5:02

Yeah.

Kristina Hoeppner: 5:03

From Mike Joiner-Hill and also Helen Chen who I had interviewed earlier in the podcast, and I'll make sure to link to their episodes as well, we learned that the Integrative Learning Portfolio Lab is part of the Career Services at Stanford and from what you've just said, also, you're focusing on digital presence and therefore more an outward facing portfolio that you can show employers and represent yourself and the work that you do. You created a LinkedIn post earlier in the year where you talk about using generative AI creating your portfolios. Why did you use that?

Mario A. Peraza: 5:40

I think the reason I used it was number one, we were already starting to talk about how we could incorporate AI into the ILPL framework that we have about, you know, digital presence curation. That was a topic that was already on our mind. But also, you know, I've been working with the Lab for about a year, and I was like, You know, I think I should revamp my portfolio with all the tools and skills that I've learned throughout teaching it and being a part of the Lab.' So that was also on my mind. I think it was winter break or something where I was like,'I'll just use AI to see how that works. Maybe it'll make my life a bit easier on recreating my portfolio' because I want to do a lot of different things to revamp it, mainly using the framework and the branding techniques that I had learned about through the Lab. That's how it started. I think I learned a lot from using AI. I definitely started by taking the ILPL framework and going through that and figuring out what my story is now, what my purpose is, who my audience is, that's where I started. From there, I started using ChatGPT. I actually have a running document of all of my experiences that I've been a part of, so my projects, my jobs, and that lives on a Google document where I can just copy and paste giant blurbs into an AI bot if I ever need to. And that is something that is super useful and super important because now I have this running list, and I can just update it. If I ever need to change to not ChatGPT or, you know, if there's a new, better version of AI, I can take that list and put it in where it can see all my data. But what it really helped with was, once it has all that and all that knowledge, it knows what your purpose is, knows who your audience is, it can reformat the data, and that's what really, really helped with streamlining my process. If I wanted my experience to be in résumé format, I could just ask Chat, 'Hey, you know, I did this for my 218 class. Can you give me this in résumé format, pulling out these type of skills?' And then it would do it. Or if you wanted a certain LinkedIn post, you could be like, 'Can you give me a summary of my experience during this job for LinkedIn?' And it would give you a LinkedIn post. For me personally, for my portfolio, I wanted to highlight my objectives, the outcomes, and my hard skills and my soft skills. So I asked it to give me, you know, different lists of hard skills, soft skills, and what you thought my objective and outcome was for the experience that I described. It did that and if you go to my website, you can actually see under all of my projects, it's formatted like that: objective, outcome, hard skills, soft skill, and that was all ChatGPT written. I will say though that you definitely have to read through everything that it's outputting because sometimes it will give you inaccurate information or just information that doesn't really sound like, you know, your voice. So I definitely had to go through everything that it was outputting and reformat it to fit who I am and what I wanted it to really express and double check to make sure that it wasn't putting different things there. But it definitely provided new thoughts, helped me brainstorm. Organisation was key, especially with the hard skills and soft skills. I think that's always a struggle for you to go through and be like, 'Oh, what did I actually learn here?' And then helping you start that process and go through and be like, 'Maybe this hard skill isn't really what I learned here. Thank you for giving me that.' It's really just a good starting tool. Obviously, not the end all be all, but it does really help streamline.

Kristina Hoeppner: 9:01

Mhh. Did you find anything surprising, then in what ChatGPT outputted to you in regards to what it thought your skills were or what you had done?

Mario A. Peraza: 9:13

Yeah, sometimes we would use very broad buzzwords, I think. I did ask it to tailor it to what I would want my profession to be. So I think it used typical things that was like, Oh, yeah, you know how to do this, this, and this.' And I'm like, 'Ah, well, that's not really specific to the project.' It was just something where I was like, 'Oh, that's a bit too broad. This doesn't really match with like what I actually did.' And that's where I'd be like, 'Maybe I'll just take these out, but leave a couple of these points in,' and then go through it again and figure out like what specifics I wanted to include, but other than that, I don't think it was very inaccurate. I think some of the times it would just be too vague or too broad or buzzwordy.

Kristina Hoeppner: 9:51

To your point of you always need to read what Chat had written for you, right before we started recording, I'd actually not Googled you, but I, I don't know how you would say ChatGPTed you? Even with the latest version [laughs], it was quite the experience because it made you out to be an accomplished professional with over 20 years of experience in higher education. And clearly you're sitting in front of me, and that would be an accomplishment in itself from very early on, in your probably still kindergarten years to have now gained 20 years of experience in higher education serving as Vice President for Campaign and Engagement at a university. And only in the third paragraph, it did mention who you actually are, but it made it read like that one person was all of you, even though it was very clearly two different people there. So yes, taking the output from ChatGPT with a grain of salt is very important there.

Mario A. Peraza: 10:53

Yeah.

Kristina Hoeppner: 10:53

You also used not just generative AI in your portfolio, but also AI to help you with the design and aesthetics of your portfolio. Can you tell me a bit about what that experience was for you?

Mario A. Peraza: 11:07

Yeah, I think actually, this was a super useful thing as well just because, I don't know, if you have experience trying to format things to look pretty. I know, I do that a lot of times, like flyers and different things like that. But there's so many different options to choose from between fonts and colours and things like that. So when I was making my portfolio, since I already had, you know, Chat open in one tab and my portfolio opened the other day, I was like, 'Okay, I want new colours, and I want it to be a cohesive colour scheme. Maybe Chat could just help with that? I really liked the colour purple. So can you give me a five-colour palette scheme where purple is the centre?' So it did it gave me a couple of different options, and I was like, 'Oh, that's cool. I like this one the best. So I'll choose this one.' And so if you go to my website, you'll see it has a very specific colour scheme, and that was helped made clear by Chat. Even the fonts on my website, I think I picked my header font, the big bold font on my website, and then I was like, 'I need a sub font that goes with that font. What are a couple options, Chat?' It gave me five different options. I looked through them and I was like, 'Oh, this one looks the best.' Yeah, I think just small details where you could spend, you know, minutes or a while going through every single font and be like,'Oh, I kind of like this, I kind of don't like this, maybe I'll keep this. 'It really helps streamline that because it's using other information from online where they have colour palettes that go together and fonts that go together, and this just helps it make it so much quicker.

Kristina Hoeppner: 12:30

I certainly see the engineer in you, everything is about efficiency there.

Mario A. Peraza: 12:35

Yeah.

Kristina Hoeppner: 12:35

And your website does look really, really nice, very cohesive.

Mario A. Peraza: 12:39

Thank you.

Kristina Hoeppner: 12:40

Oftentimes, that's something that one doesn't really notice as much, I find, but you notice that the design is off. And so that's really great to see how you use it. So you didn't use a tool in the platform that you were using at that time, but asking ChatGPT in order to come up with a colour palette, and then also a matching font. That's really cool use..

Mario A. Peraza: 13:03

Yeah, thank you. I know, I wish I had a screenshot between my pre AI portfolio to now because it's such a difference [laughs]. So clean now.

Kristina Hoeppner: 13:12

Maybe there's a history function somewhere on the website generator that you're using or maybe even the Archive has it archived. That definitely would be cool to see. And also then show in your workshops with the students so that they can see the difference in what they can do because often we don't know the questions that we should be asking to get certainly answers. Having your experience in there telling students what you have done, the results you have achieved, and that you like the efficiency, that it was important to you because of course you can scroll through 10,000 pages of colour palettes, but it does take some time. And so when you hear those examples, I think that's really, really helpful to use with the students and then help them get similar results. Now speaking of the workshops actually that you are doing, you created a new workshop in which one you use AI to curate your digital presence. You've actually only run it last week. So I'm curious in finding out how that workshop meant and what the students reactions were there.

Mario A. Peraza: 14:16

Yeah, yeah, I know, this workshop has been talked about for a while, but I think me actually, you know, going through the process myself helped get it to where it was because, you know, it's a very new topic, and I learned so much using AI myself to make my portfolio, but also I learned so much trying to create this workshop. I was going through the flow, I obviously want to use the ILPL framework because I think that framework really helped me with the entire portfolio making, LinkedIn making, full digital presence curating there. When I was going through 'How should I format this workshop? How should I format the slides?' I went through each step of the framework, and I was like,'Where could AI be used?' What we ended up realising was that AI can be used in every step of the framework. And really, I think something that I noticed when I was using tools was that if I just had guiding questions or guiding prompts, how I could ask ChatGPT to figure out who my audience is, to help me with my purpose, to help me brand myself, then that would have been so much easier along the way. What we did is we actually made this worksheet. It's like a two-page worksheet where for each step of the framework, and there's six steps, there's a couple of guiding questions, and then a couple of tips and tricks on how to make this easier, how to make this more accurate to what you're looking for. So for the very first step, it's'Search yourself: What is your story now?' you mentioned searching my name, it kind of gives you different options there. That's what we've had the students, too, during the workshop. We were like, 'Okay, let's use Copilot because it uses the Bing search engine, and search yourself and see what it gives you. It even gives you the reference of where it came from. During the live demo, I searched my name, Mario Peraza. Obviously, it gave me five different Mario Perazas in the world, only one of them was accurate to who I was. But then I searched Mario A. Peraza, and it gave me much more accurate information because I, you know, built my brand around, including my middle initial, which is something that we also talked about in the Lab. If you have a very generic name or multiple people use this name, you should have something that makes you stand out, and so having that middle initial pulls it more directly from my website, just because my website is Mario A. Peraza. And so it'll show you where it's linked to. I thought that was super cool that it could pull the information and you can see what your story is now in a cohesive way where it's like, 'Oh, this isn't really what I want it to be giving information about' or 'Yeah, this is exactly accurate. This is what I wanted it to be telling me.' I think going through these examples that we were showing live beforehand, I was like, 'Oh, wow, this is actually a super useful tool along every single step of the framework. How can we make this easier for students to use?' So we made this, we sent out the document during the workshop so that people follow along with us as we went through each example. And one of the students actually said, and this is a quote from them, "I believe the prompts enabled me to come up with my own questions to tailor it to my own needs. It's an extremely useful way to get started." That made students feel less overwhelmed and made students feel like they could jump over this hurdle of getting started because when you're redesigning your portfolio or your LinkedIn, it seems like it's just like a lot. You have to remember all of your past experiences, but what we really want show was that this was a great tool to get over that initial barrier. It's a lot of front work, you know, you have to provide it with the data that you want it to know, you have to provide it with how you want to tailor your experiences and things like that. But once you have all that, if you save it in some sort of token or this running document that I talked about earlier, you can always use it, and you can get information reformatted in any way that you want. Everyone's growing, and everyone's changing. So maybe my purpose now isn't going to be what my purpose is in the future. But if I already have all this data saved, and I know what my new purpose is, it can retailor it without you having to go back and think, 'Oh, how can I frame this in a new way?' I just think that that's the beauty of it. And yeah, that's what we heard from the students as well, it was it was such a great workshop.

Kristina Hoeppner: 18:14

Oh certainly sounds like a really, really good one, especially also seeing what the internet currently knows about students who might not always have a very huge presence and therefore the likelihood of the AI just making something up can be higher versus then also seeing what the internet knows about other people and then seeing how you can adjust the prompts in order to get better responses.

Mario A. Peraza: 18:39

Yeah.

Kristina Hoeppner: 18:40

I really also like that you're still emphasising that is part of the brainstorming, part of getting over maybe blank page syndrome, where you don't really know where to start, that you help people find their voice, see, do I want to say it this way? Do I not want to say it this way? Do I want to include this experience or rather different ones, rather than just copying and pasting what the AI spits out. You also just mentioned, Mario, that, of course, especially when the AI does know a lot about you, you do need to feed it with all of that information of what you have done, when you have done it, what you've learned, and all of that. You do need to share then quite a lot of personal data actually that might then be used by the AI to train future models and data that also then flows into a general data set. Was this a concern when you recreated your portfolio? Or was it also concerned when you had the conversations with the students?

Mario A. Peraza: 19:40

There's been a lot of talk amongst Lab members of this point that AI is a very new tool. You know, data privacy is also a huge topic and big issue. We made sure to include an explicit slide in the workshop that was like Don't share any super personal information as secure as people make programs and platforms, nothing is truly secure online. So definitely don't be sharing passwords, Social Security's, you know, things like that.' And you know, informing them that each AI and each company has its own different policies. And just to be aware of those and to know that your information can be used for research or improving AI tools. So making sure that was clear and open was a big point of ours, especially during the workshop and letting them know, obviously, AI is an optional tool, you don't have to use it, especially if you don't feel comfortable, but it is here, if you'd like it. I think it's still a big concern, a big topic. But for me, personally, I knew when I was using this tool that I wasn't going to be giving it anything that I didn't want the general public to know because it is portfolio creation, and I'm putting it on a publicly accessible website. So I didn't include anything personal. I kept it broad, which is also what I told the students to do,'Keep it broad, don't do anything too specific or personal to yourself.' Yeah, like I said, it's the information that I wanted to be access to the general public any way. I think using the tool to search myself and seeing that it's giving me the results that I really tailored the internet to find just shows that it actually is working. So yeah, that's that's where that is.

Kristina Hoeppner: 21:16

I think that's where there is the advantage with your portfolio that you created and also the portfolio's that you support students with in the ILPL is that you are looking at the digital presence and creating a publicly accessible portfolio because then as you said, you're already more aware of, 'Hey, I'm only wanting to share things that everybody can know, that I'm happy to share publicly.' Whereas if we are thinking about portfolios for nurses or for teachers where they might also need to share photos or want to share photos from their students or really experiences that shouldn't make it public, then using AI for that might have quite different implications and different concerns then, right?

Mario A. Peraza: 22:01

Yeah.

Kristina Hoeppner: 22:01

You were using Copilot during your workshop and experimenting with it also in the ILPL. Do you know if the university then actually vetted it because that is part of the Digital Ethics Principles that we've created in the AAEEBL Digital Ethics Task Force that when you use software and when you ask students to use an electronic tool in particular, that they should be feeling comfortable using it and that the university should actually look into the data privacy and ensure that student data wouldn't be compromised rather than just saying, 'Students, here's a tool, read the Privacy Policy and Terms and Conditions and then make up your mind' because of course, if it is required for a class or should be used, then the university, I think, does signal in a way that they are okay with using a particular tool. Did you do any of that work beforehand as a team?

Mario A. Peraza: 22:51

I don't know if Stanford does particular vetting on different AI tools, but we did make sure that we said it was completely optional, and the reason that we use Copilot was because it uses the Bing search engine rather than ChatGPT. But we only used Copilot for the very first example, which was the 'Search yourself, like what is your digital presence now?' Typically, for that example, you usually partner up and search your partner on the internet. We asked them to open a private tab and see if you can find information about your partner on the internet. But for this workshop, we felt like using an AI tool would be more appropriate for that. It was completely optional, and you didn't have to do it, you could also just use Bing instead if you didn't want to use Copilot because it's the same search engine, it just wouldn't make all the information succinct for you. But for the rest of it, we did use ChatGPT as the main example. But it wasn't about searching yourself, it was more about the running document that I was telling you about. You can put this data in if you want and then it will reformat it to whatever you want. But I do know there's a bunch of different AI initiatives and clubs and councils on campus right now, and I think there's a bunch talking about how it can be used in different classroom settings, but that's all still very new. And so we want to make sure that we express that all these tools are definitely optional and don't have to provide any information that you don't want to.

Kristina Hoeppner: 24:16

Right. Thanks. Are you also planning on creating other workshops that look at the intersection between AI and portfolio and data, data privacy and the like?

Mario A. Peraza: 24:27

Since I'm graduating in a couple of days, I will unfortunately no longer be a student liaison at ILPL, but I do know that they are very interested in this, and they really like that we have a starting place for this because it's going to grow and expand. You know we have this initial document with all these prompts and things, and so I think, there's definitely more work to be done here. There are going to be future workshops through ILPL, and I'm super excited to see what they kind of do with this.

Kristina Hoeppner: 24:59

Are you planning on keeping your portfolio updated, your portfolio website?

Mario A. Peraza: 25:03

Oh yes, of course. Like I said, now that I have that running document, I'm just going to keep adding, I kind of have to do it this summer because I've had a variety of different experiences that are not yet on my portfolio. So I'm going to do that. And I'll know even better how to incorporate AI to make it a little bit easier for myself.

Kristina Hoeppner: 25:21

Yeah. Now with the tools that you're using and also AI, is there though anything that you'd like to be able to do with portfolios, but currently can't?

Mario A. Peraza: 25:31

Yeah, I think something that would be super interesting, which also might be centred around AI, but I didn't even think about it that way. There's always this part of the workshop where we talked about the difference between in person and online presence. If you're in person, you know who your audience is, and you can give answers tailored to what they would want to hear or relevant information you should share with them. Whereas online, you search a website, and then you go through it, and then that's all the information you're getting. I think if there was some sort of bridge between those two areas where maybe whoever is logged into your website, there is some sort of helpful guiding bot or something that could be like, 'Oh hello! What would you like to learn about Mario Peraza today?' And then it could just tell you is like, 'Who are you? What would you like to learn? Why are you looking at his website?' You know, something like that. That would be really cool. I mean, as my portfolio grows, it's gonna be harder for people to actually look through everything as well. I think I already have a ton of projects on it, where it would take someone a while to actually go through it all. But if there was, you know, something that was just like, 'Hey, what would you like to learn? This might be interesting for you to click on. There's this page on his website that could be interesting to click on.' So yeah, I think that would be, that'd be kind of cool.

Kristina Hoeppner: 26:43

Yeah, because right now, people need to create separate portfolios in order to tailor to different audiences or to tailor to different employment opportunities and the like, and you would just automate that, in a way in true engineering fashion, automate as much as possible there in order to also do that more dynamically rather than needing to update multiple portfolios there. That's a really cool idea. Thanks so much for sharing that, Mario. Now to our quick answer round. Are you ready?

Mario A. Peraza: 27:11

Yes.

Kristina Hoeppner: 27:12

Which words do you use to describe portfolio work?

Mario A. Peraza: 27:16

I would use reflective, expressive, and intentional.

Kristina Hoeppner: 27:21

Awesome, thank you. What tip do you have for learning designers or instructors who create portfolio activities for students?

Mario A. Peraza: 27:30

I think just having hands-on examples where students can follow along that has always been the best in our workshops, and Google yourself. Everyone loves that [laughs]. So that's been a great one.

Kristina Hoeppner: 27:44

What advice now do you have for portfolio authors, for the students creating portfolios?

Mario A. Peraza: 27:51

Definitely spend some time doing the reflective work beforehand and really knowing who you are and what you want others to know about you. That makes a big difference in the end.

Kristina Hoeppner: 28:02

Mhh. Thank you so much, Mario, for sharing how you have been using AI in your own portfolio experience, as well as in a workshop with students that you've just facilitated. It was so good to have you on the podcast.

Mario A. Peraza: 28:15

Yeah. Thank you for having me.

Kristina Hoeppner: 28:16

I wish you all the best now for your professional career wherever it might lead you after graduation. And yes, looking forward to seeing your portfolio and how it evolves over time.

Mario A. Peraza: 28:28

Thank you.

Kristina Hoeppner: 28:30

Now over to our listeners. What do you want to try in your own portfolio practice? This was 'Create. Share. Engage.' with Mario A. Peraza. Head to our website podcast.mahara.org where you can find resources and the transcript for this episode. This podcast is produced by Catalyst IT, and I'm your host Kristina Hoeppner, Project Lead and Product Manager of the portfolio platform Mahara. Our next episode will air in two weeks. I hope you'll listen again and to tell a colleague about our podcast so they can subscribe. Until then, create, share, and engage.