Guest: Kayvon Touran

In this episode, Dr. Robin Sargent interviews Kayvon Touran, the CEO and co-founder of Zal.ai, about the role of generative AI in workforce learning and development.

Tune in to find out:

• How Kayvon and his company are using generative AI to enhance employee training and practice through AI-powered practice partners and feedback facilitators.
• The importance of using quantitative and binary rubrics to measure employee skills and competencies and how this data can drive personalized learning interventions.
• Kayvon's advice for aspiring instructional designers on embracing AI as a tool to augment their abilities and play a crucial role in the future of workforce development.

Listen to this episode below: 

Subscribe to Become an IDOL Podcast: Apple Podcast | Spotify | Google Podcast 

Connect with Kayvon on LinkedIn

Are you looking for a no-nonsense formula for creating engaging courses and training? Check out my new book, The Do It Messy Approach: A Step-by-Step Guide for Instructional Designers and Online Learners (IDOLs) on Amazon.

Enjoy the Episode Transcript below:   

Dr. Robin Sargent 

Welcome to Become an IDOL. I'm Dr. Robin Sargent, owner of IDOL courses. This is the place where newbies come to learn and veterans share their knowledge. I have here with me today, Kayvon Touran and Kayvon, would you please introduce yourself?

Kayvon Touran 

Yeah, my name is Kayvon Touran. I'm the CEO and co-founder of Zal.ai.

Dr. Robin Sargent 

Wonderful. So tell us your background. What does your company do, first of all, and then I want to hear your background. How did you get into this space? Where did you start? But tell us what your company does.

Kayvon Touran 

So we help companies and L&D leaders at those organizations contextualize, measure and apply their learning and development for their workforce. And we sit sort of at this interesting intersection of generative artificial intelligence, learning science and workforce development, and are producing products that sort of sit at that middle ground. So that's what my company does, and I can talk, you know, we'll get into it more, but at a high level, that's what we do. And, it's been an interesting journey, sort of getting here. I've spent my entire career working in early-stage technology startups, and I spent the last six of those years working at an EdTech incubator in New York City. I worked at this company called Noodle that was founded by John Katzman. He founded the Princeton Review, as well as another company, called 2U. And while I was there, I sort of got a lot of different experiences working on products that were helping everything from sort of pre-K to post-secondary educators. But the most recent project that I was working on was a lifelong learning platform, and so we were trying to build a competitor to Coursera and edX. And the sort of interesting insight that I had when I was working on that project was we would go to these schools of continuing education and professional study, and they tell us, it's interesting we'll have companies approach us to develop custom learning, and they're not really sure which of their employees need learning. They're not really sure how they'll be able to apply this learning once they go through the sort of training program, and they definitely don't know how they're going to measure the efficacy of the investment they're making with us. And that sort of struck me as really fascinating. I mean, I kept hearing this from all these different really reputable schools that that was sort of the dynamic with companies. And so, I took a step back and realized that there's likely a lot of value that we could help create for those organizations and really for the end users, the learners themselves, to help them really contextualize that learning, be able to apply it more effectively, and also to allow individuals to be able to really measure their development, because it's such a big part of how our careers are sort of evolving now, for a lot of reasons that we can talk about. So, that was sort of the origin story. And from there, one thing led to another, and I quit my job and started Zal with my co-founders.

Dr. Robin Sargent 

So, you've always been in the early tech space, and then you got into the EdTech and so that's kind of how you found, like this workforce development side of things, and then you basically found a unique problem that you could solve with your skills. So what did you actually major in in school?

Kayvon Touran 

It's a great question. I majored in business.

Dr. Robin Sargent 

Oh, okay.

Kayvon Touran 

Yes.

Dr. Robin Sargent 

Okay, and so now, when you and your partner, I'm sure that now you're all in, right? Because when you're looking to solve a problem like workforce learning and then also making sure to measure and see that is effective and things like that, I'm sure that you've done a lot of deep dives. So, what is kind of your favorite topic right now? What is the thing that you are really interested in, really seeing some gains as far as what you and your company are doing?

Kayvon Touran 

You know, one of the topics I think, is really interesting that we don't talk enough about in workforce development, at least I haven't seen, is how we make the connection to practice. And I think there's sort of this habit that I've seen when it comes to workforce development, where there's almost the separation between the educators in the organization. So those that might be the L&D professional, that might be the sales enablement professional, they're creating content, and they're trying to help their workforce learn new concepts. But then typically, and I'm talking about sort of like a professional or like a knowledge worker setting, typically, someone takes, let's say, a course, and then it's sort of up to them to really apply that on the job for the first time, a lot of times. Or maybe there's, like, some light role-playing scenario-based stuff that they do in person and training, which is great, but what I think is a massive potential is, how can we help those employees practice and rehearse these skills? And it's interesting that that isn't more common, right? Like, think of an athlete, think of a musician. We were talking before we jumped on, what is trumpet that you used to play?

Dr. Robin Sargent 

Yeah.

Kayvon Touran 

So, imagine if you had to show up to the concert, and you'd only read books and watched videos on how to play the trumpet, but you hadn't spent all that time practicing by yourself and with your bandmates. I mean, I feel workforce development, in some ways, is kind of similar a lot of times. And so I think there's an enormous potential to help companies and their employees actually practice skills before having to actually apply them, but also in sort of a more iterative process of going to apply them at work and then practice what they've learned. Yeah, so.

Dr. Robin Sargent 

It sounds like I'm talking to myself right now. I am obsessed with deliberate and applied practice. It is something that I've been preaching for a very long time. I am really excited to hear that we share exactly the same perspective that a lot of people are going around blowing the wrong notes with no lip muscles because they don't practice, so to speak, just to keep that analogy going. But so how does, and I love to deliver practice, I think, just like you said, you can't, it's different to just read music and actually play music, those are two very different things. So how does gen, like generative AI, come into play as far as enhancing the opportunities that employees and workers have to do more of this, hands-on, practice, and applied training?

Kayvon Touran 

Yeah, it's a great question. So I see generative AI kind of playing two roles in this process. One is that it's your practice partner, so it can act as that external customer or internal team member, and it can help you actually apply and practice really any durable skill or competency, right? And that's something that's really big for us because it's kind of this paradox: as technology continues to advance at an increasing rate, the half-life for all of our heart skills is getting shorter, and the importance of our ability to be agile and be adaptable in the workplace. And a lot of the durable skills that are part of that are becoming even more and more important. And so being able to practice those things, I think, is incredibly important. So specifically, the AI can help you be that practice partner, but it can also be your feedback facilitator. And I can talk more about sort of how that works, if you're interested.

Dr. Robin Sargent 

Of course. Now, you know, I have ideas about how you would do that, but I imagine so we have a lot of people who are new to the industry that listen to this podcast, and so they're probably thinking like, okay, well, that sounds complicated, or it sounds difficult, or maybe they've never used generative AI in that way. So maybe we could get into a little bit of the specifics about how, even someone new to this field could start thinking about leveraging generative AI that way. Are you talking about building your own GPTs, or what is it and how does that look?

Kayvon Touran 

That's actually so that's a fantastic question and the reason why I like that so much is because I think with a subscription, not even a lot of times with subscription to one of these platforms, whether it's chat, GPT or whatever, you can experiment with some of these approaches and sort of see the potential for it. So yes, how would you practice something like this yourself as an L&D professional? Well, I would ask that an L&D professional, go into ChatGPT or something and prompt it around this idea of, hi, I want to create a simulation, a role-playing scenario you're going to act as, let's say, an internal team member, part of the project management team, and I'm going to pretend to be myself. And what I want to do is practice my ability to explain to you how the L&D process works, or whatever, right think of scenarios that are really specific to your job, that are that are probably pretty tough right for you to train on. I know, part of what makes being an L&D professional so hard sometimes is explaining to other departments, as far as like, how your process works and making sure that they can understand sort of how you fit into that process, how they fit into the process as a subject matter expert, so taking like a scenario like that and training around it, I think you can start to see just by doing that and going through that exercise and then asking the AI, how well you did, I think you'll see the potential. Now that is just sort of like a drop in the bucket. What we're doing and what I think it kind of takes it to the next level is okay, now, how would we consider the specific rubrics of proficiency that you've probably developed? It could be anything from a nine-box and something role-based to something very specific around a particular competency or skill and what it looks like to be proficient at that. Now what if we trained our AI on that rubric, and we also train the AI on things like a job description and other information that makes it really specific to the types of experiences that one of your training constituents is going to experience. So now the people who you're helping at the organization are practicing scenarios that are hyper-realistic to their job, but also getting feedback on the sort of, you know, behaviors, expectations, and proficiency expectations that you've set for the organization. And then now we can track all of that, and we've built this whole system around it, right? That's where I think things get really, really interesting. But to your point and to your question, it's so easy to see the potential of this if you just start kind of experimenting with something like a ChatGPT. I really do think it captures a lot of people's imagination when they start doing that.

Dr. Robin Sargent 

100%, I have built so many little GPTs. And just to follow up on what you said is that you don't even have to have a paid subscription. There are some other tools out there, like Party Rock, for example, where you can build your own custom bot for free. You don't need a subscription. And so if somebody wanted to go and start playing with that and doing exactly what you said, where you upload the rubrics and you put in some of the rules and things like that, then, I mean, who knows how sophisticated you could make it, but you could get a really good start as far as, like, putting some of those practices into action. So when you talk about, like, what your company does, like the role play assessments, is it just tech space?

Kayvon Touran 

No, definitely not. I mean, I think, so some of our early pilot partners are actually in professional services, PR firms, law firms, and that's one of the great things I think, about an emphasis on durable skills, is that it's such a big part of any job. But also in thinking about having a system that's adaptable enough that can really be based off of any rubric of proficiency, that layer of contextualization, right? Because that's my I guess this is one of my, one of my gripes with generic skills-based education, by the way, I think it's way better than not having it. But even thinking of something like conflict resolution and training on conflict resolution and developing a generic course on conflict resolution, I mean, we as humans are really bad at sort of that far transfer of saying, okay, I've learned this general concept of conflict resolution, and now let me just go about my day and apply it everywhere I see. That's really hard. So, our ability to sort of contextualize that very specifically, not just to a company, but to a role and how that person might experience it, I think, is a really big unlock. And for us, we can be, we can sort of work with anyone who is in the workplace, because those skills apply to everyone.

Dr. Robin Sargent 

I think I can hear people, if I was to guess they would like their next question would be something like, okay, so this is cool. Maybe I have experimented with some of these tools to be able to do some of these things that you're talking about Kayvon. So, what does your analysis process look like? Does it differ at all from kind of the typical process that they might go through as an instructional designer in the workplace? Do you do anything extra to make sure that you are creating these rubrics or gathering these documents, or setting up the competencies or the prompts in a certain way that gets you better results?

Kayvon Touran 

Yes, absolutely, that's such a great question. So my co-founder is a learning scientist, and he thinks of the rubrics like one of the fundamental things that he teaches, and obviously any company that works with us gets access to his consulting firm, and we help them define some of these rubrics. But one of the tenets that I can share is moving away from qualitative rubrics to more quantitative ones. And emphasizing the importance of sort of binary definitions, right? So, we can't say proficient looks like someone who does this really well. Well, that's going to be really hard for anyone to interpret and analyze, and that's a problem with a lot of these rubrics. It's going to be a problem for an AI, for sure, and we don't want the AI to be reasoning in that capacity, but it's definitely also going to be a problem for any human because now you're leaving it up to every manager at the organization, every person to sort of define what really means, right? So we can't, we have to move away from some of that more fluffy language, and that's hard, but by being more binary and sort of saying it looks like this or it doesn't look like that, we can then actually use a sort of more clear definition or analysis on did this person do this thing, or did they not? Did this person do that thing or did they not? And sometimes, that analysis can't just come from a self-assessment. It has to come from the experiences of your colleagues and, yeah, and your teammates who are sort of working around you as well, and that sort of ties into some of the stuff that we're doing as well.

Dr. Robin Sargent 

Man, you just gave me so many good ideas because we use a lot of qualitative rubrics in my vocational school, and it is exactly what you say where, like, one evaluator will interpret it this way, another evaluator interprets this way. And then, like, the students might get a different evaluator to, like, return the asset, and then they're frustrated because, well, this evaluator said it was fine, and now this evaluator says it's not, and the idea of just moving to a quantitative rubric is like, oh yeah. Oh, that makes sense.

IDOL Academy Virtual Open House Ad

If you're listening to this podcast, you know this is the time to start working smarter and not harder towards your IDOL goals. Join me on August 15, at 6 pm Eastern Standard Time to learn more about our proven and state-authorized Career School IDOL Academy. You'll learn about how our school is a career accelerator in the field of instructional design and online learning development and why 100% of our graduates land prestigious positions in learning and development within six months of graduation. To register, visit idolcourses.com/events click on the virtual open house sign-up button; and I look forward to seeing you there.

Dr. Robin Sargent 

And so I'm just wondering, so your rubrics must have, I don't know, 100 items, whereas, like a qualitative one, you got 10, right? And then you got a range, usually. Is that kind of what it looks like when you talk about your quantitative ones, you got, like, many more items, or do you just get more, do you narrow down what you think is important, yes or no? Like, as far as your binaries go.

Kayvon Touran 

More the latter. And you know, the rubrics can get pretty gnarly if we want them to. But as a starting point where we would suggest we start with an organization is sort of simplifying them, and to that point, simplifying is hard. Simplify does not equate to easy. I actually think that in a lot of cases, we get qualitative because it's easier. Because it's easier to just to say, yeah, you really got to do a good job at that. It's like, okay, that made my job easier as the person creating the rubric, but you've made my job harder as the person analyzing it and evaluating it.

Dr. Robin Sargent 

Yes, yeah and even, there might even be some benefit to where, if you make the quantitative and you program that into the generative AI to analyze first, and then it would flag things to look at and then maybe you could have more of a qualitative even after you have, like, an AI intervention that can flag, hey, they didn't do this, so they didn't do that, because sometimes some of it just takes a human eye, right? I'm thinking, of course, in our field, right? Like an e-learning course, right? If somebody creates an e-learning course, the AI is not going to be able to look exactly at everything, like, just like somebody who's already a professional in the field, like, see other things, so it still needs a human eye to look at that.

Kayvon Touran 

Absolutely, yeah, and the way that we think, I mean, the way that we think of that isn't just your eye, it's also the eye of your colleagues, right? So traditionally, 360 interviews or reviews is seen as like a great way to measure durable skill competency. And so we think of an almost this combination. If we are able to create a more quantitative rubric and someone goes through a self-assessment, we can have an AI, do an analysis of that and identify here's what they did well, here's what they did poorly, here's where there's room for improvement. And obviously we serve that up to someone in instructional design or a manager or someone to consider that before it's then shared with the end user, because we need that last evaluation and human touch in the process. But then we also have our AI interview, your colleagues, your supervisor, etc., and it's incredible. It uses the same rubric and it asks the same questions, or it asks a series of questions to kind of extract, okay, over the last quarter, or since we did this last how has Robin been demonstrating conflict resolution? Give me a real example of that. And so we're collecting all this qualitative information about that as well, and then we package that all together for the manager, again, instructional designer, what have you, where they can see not only the original transcript of the self-assessment if they want to dig in deeper, but they get almost like, I don't know if you remember those restaurant reviews, the Zagat style restaurant reviews, where it has all the quotations of what people thought about. So Robin is "energetic and, etc., etc.". And so, yeah, there's just so many interesting ways, I think, to leverage this technology to get deeper, richer information, but also do it in a way that is scalable, that is kind of the equivalent of us just sharing around a survey, but getting so much richer information out of it. And then as instructional designers, obviously having that rich data can lead to, I think, so much more impactful learning for the organization, and one that's baked in data. And I think that's a really interesting and big shift for L&D and the sort of AI age.

Dr. Robin Sargent 

That's one of the things that I love because now I'm also thinking, well, not only are you collecting all this data around the rubrics and their abilities and then what they are saying about, you know, these employees on the job, but if it's all going through generative AI, you hook that up to like an automation tool-

Kayvon Touran 

Yeah.

Dr. Robin Sargent 

Or you build your own automation, which I imagine a company like yours would do. I am just a Zapier fan. I just Zap everything.

Kayvon Touran 

Zapier's great.

Dr. Robin Sargent 

But now you could take all this data and have it, look at it and say, okay, what are these, what's the areas that they keep getting wrong, or what are the places that they all seem to struggle with? And then maybe that's another intervention that is created, or you can do maybe a bigger analysis across because I'm sure, like, a lot of the competencies go together, and so you look at more of a whole picture, and then you can create other tailored interventions, because that's like, kind of what we were talking about to get started, right? It is not just skills-based, but skills in context.

Kayvon Touran 

Yes, and you have a crystal ball, because that's how exactly right. So with all of that output comes really rich reporting for the L&D team, where they can look at things at a company level to a department level, to a team level, to a team level, to an individual level, and sort of identify where those interventions would be most impactful, and then, crucially right, we can take them through another round of assessment if we want, and now we can actually show, look, this was a place where people weren't super clear on and then they went through this intervention, and they went through a similar role-playing scenario, and look now they're doing really well in that and I helped create that movement, and it's all baked in numbers. And so we're speaking the language of the boardroom, so to speak, as L&D folks. And I think that's really powerful in the direction that the industry wants to go in as well because it just helps them sort of prove all the value they're creating in the workplace.

Dr. Robin Sargent 

When you know what the competencies are and you have all these measurement tools in place—pre-measurement, during, after, and so on—you can also influence the hiring decisions that they're making. I mean, as we all know, like businesses, their best asset is their people.

Kayvon Touran 

Yeah.

Dr. Robin Sargent 

And so to be able to even get better people on their team that have the natural aptitudes for some of these competencies as well, and you now know how to measure it, because you've already done it through, have you found anything like that, then using it now?

Kayvon Touran 

So yeah, again, you've got a crystal ball. I see the opportunity for organizations starting internally. You know, I think there are a lot of companies that are thinking about these ideas externally for recruiting purposes, but I actually think there's a lot of value for organizations to start internally with their own workforce because you have so much better access to information and communication and transparency with your own workforce, and what you can do is really refine essentially what proficient looks like for and identify the competency catalog and what proficient looks like, and how you look for those things and that, and work that into the process. And then there's an opportunity to sort of externally, sort of express that potentially, right is that we could look at how we then recruit based on those things, and it could either be based off of the data that we're collecting internally, to say, look, here are some skill gaps and competency gaps that we could really use some external hire to come fill, because we just don't have it internally, but also in roles where they're really well defined, we have a really good sense in terms of interviewing for those roles on what to look for and how to measure them. And we could arm all of the people doing the interviews with that information as well. And so I think it gets really, really exciting and interesting as it sort of evolves into thinking about sort of externally too how we bring people into the organization. But in my opinion, I think it's a mistake for companies to start externally. I think there's a lot of value in starting internally, first, starting small, and then really creating that into a competitive advantage for the organization.

Dr. Robin Sargent 

Yeah. So, I think the other thing I would ask is, okay, so you load up generative AI, you've got all your quantitative rubrics, you've got the company policies or whatever kind of handbooks that you can load in there. I'm just wondering, what does testing look like? How do you know it's ready to roll out? What do you guys do around that?

Kayvon Touran 

That's a great question. So, my recommendation for how we roll out every time, and usually companies are really excited by this is, I say I want your harshest critic first. And so typically when we engage in an initial paid pilot, it's with the L&D team themselves. So I want them to go through this experience, and I want them to see sort of the potential for it. I know that they're going to be the harshest critics about everything that it's doing. And then, you know, typically, if we can win them over, then it sort of opens up the conversation and thinking of, okay, how could this be applied to the rest of the organization? And, you know, it starts with, like, let's bring in one subject matter expert. You know, it's not so different from, I think, the instructional design processes in any other sort of modality or format; it's similar. It's just the way that it's delivered is different. And, you know, it's not necessarily- I want to also really emphasize this. It doesn't necessarily mean that, like we're getting rid of any other content modality, either. The idea is that it sort of sits in between. It's this idea of taking all of that education, foundational learning that you're doing, and then bridging the gap between actually doing it on the job. Is that sort of application and that contextualization, that application, and then that sort of measurement layer, that's sort of where we're sitting. So, so yeah.

Dr. Robin Sargent 

I can see how that could work very well too, right? Just like you said, you're not gonna throw away everything that they've already created, but you are putting in the role-play the applications, right? The thing that can collect all this really rich data and then, you know, keep moving it into other places, to keep enhancing all the other things that they are doing, or even, like groupings of the content that they have probably.

Kayvon Touran 

Yeah, exactly.

Dr. Robin Sargent 

Yeah. Well, man. I mean, I feel I love doing these podcast interviews Kayvon because not only, of course, the sort of the people that are listening, but I got a great idea today about kind of looking at quantitative rubrics instead of those qualitatives. You know, I just, I knew that we were having trouble with but I just didn't, I just thought we just kept being more specific in the qualitative way. But it makes so much more sense to just yes or no. Like, just make it binary.

Kayvon Touran 

Mhmm.

Dr. Robin Sargent 

Genius. Sometimes it's those small shifts, right? It's like those like, 1% shift that just makes all the difference. And so my final question to you, Kayvon, is, what would be your best advice for those who are looking to become an IDOL, an instructional designer during this age, during this tech space, like based on all the things that you learned and all the ones that you've worked with, what is it that you would tell those people that are looking to break into this industry?

Kayvon Touran 

I would embrace and obviously this is going to come from me being very bullish on technology, but I would embrace this sort of paradigm shift with gen AI. I've been shocked in speaking to a bunch of people that a lot of people are just discouraged or intimidated to even experiment and play around with a prompt. And like you said, there are free resources out there, and it's not ever going to replace us. It's going to be a tool to help augment our ability to do things and being the individual, I'll get off my soapbox in a second, but being the individual at an organization that's doing instructional design that can embrace this technology is going to be so important over the next five years. And here's why, because every single job at pretty much any company that's baked in knowledge work, whether that's professional services or tech or whatever, almost every single job is going to change. The skills that you're going to need are going to change, and we're not going to be able to just hire a bunch of AI engineers for every job that wouldn't even work. What we're going to need to do is train that marketer, that project manager, that salesperson, on how to do their job, plus AI and who's going to be able to train them. I think L&D is and I think L&D is going to have such an exciting role in helping make that transition. And it's, it's really very similar to, like, the introduction of the web, and how we do all of our jobs now, I'm able to zoom with you and have a podcast, and all these amazing things that were created by the web. That same shift is going to happen with AI and L&D is at the center of it. They're going to have the amazing role of helping the company move into that age. And so I think embracing that idea and pushing on that, I think you have a lot of exciting there's a big, bright, exciting future for L&D in this world. And so that, that's what I would say to someone who's trying to break into the industry. So.

Dr. Robin Sargent 

I love it. That's great advice. And you're so right, all you have to do is approach it with, like, childlike wonder. You really are not going to break it, like, just do whatever. Like, it's not going to explode. It's not going to, like, take over your computer or anything like that. Just go and play and learn it like a child would. And I just have really enjoyed our conversation Kayvon. So where could people find you? Where can they connect with you, and where can they learn more about your company?

Kayvon Touran 

Yeah, I've really enjoyed it as well. So you can find us at Zal.ai. I'm also on LinkedIn, and always happy and excited to connect with anyone in this industry, and I'm just excited to speak with more people about it.

Dr. Robin Sargent 

Wonderful, thank you. Thank you so much for listening. You can find the show notes for this episode at idolcourses.com. If you like this podcast and you want to become an instructional designer an online learning developer, join me in the IDOL courses Academy, where you'll learn to build all the assets you need to land your first instructional design job, early access to this podcast, tutorials for how to use the e-learning, authoring tools, templates for everything course building and paid instructional design experience opportunities go to idolcourses.com/academy and enroll or get on the waitlist. Now, get out there and build transcendent courses.

If you are a new listener to Become an IDOL, we would love to hear from you.  Please visit our Contact Page and let us know how we can help you today!

Subscribe & Review in iTunes

Are you subscribed to my podcast? If you’re not, I want to encourage you to do that today. I don’t want you to miss an episode. I'll be publishing a new episode every two weeks and if you’re not subscribed there’s a good chance you’ll miss out on those. Click here to subscribe in iTunes!

Now if you’re feeling extra loving, I would be really grateful if you left me a review over on iTunes, too. Those reviews help other people find my podcast and they’re also fun for me to go in and read. Just click here to review, select “Ratings and Reviews” and “Write a Review” and let me know what your favorite part of the podcast is. Thank you!

Want more tips on how to become an IDOL? Get my free guide here

Join the Conversation!

We have a wonderful and friendly community of expert and novice IDOLs in the Become an IDOL Facebook Group. We would love to have you join us in the group or chat with IDOL courses on of our social media channels.