In this podcast episode, Dr. Beverly Wright, Vice President of Data Science & AI at Wavicle Data Solutions, and Geneva Taylor, Director of Data Analytics and Insights at Shapiro+Raj, discuss the evolving role of AI in businesses of all sizes. They explore how AI is no longer just for tech giants but also for small and medium-sized businesses looking for accessible ways to use AI. This conversation sheds light on how AI is transforming business operations and solving problems for companies across industries, regardless of size.
Speaker details:
Watch the full podcast here or keep scrolling to read a transcript of the discussion between Beverly and Geneva:
Beverly: Hello, I’m Dr. Beverly Wright, and welcome to TAG Data Talk. With us today, we have Geneva Taylor, and we’re talking about AI for the rest of us. Geneva is Director of Data Analytics and Insights at Shapiro+Raj. Thanks for being here.
Geneva: Thank you for having me.
Beverly: I love this title. Let’s start off with a little bit of background. Tell us, why are you so cool?
Geneva: For starters, and I may be slightly biased, but this hair.
Beverly: Yeah, it’s pretty cool. I’m not going to lie. You get a lot of comments, all the time?
Geneva: It depends on the quality of the hair day. Today was a good day. But outside of that, what makes me cool is that I like studying how people do what they do and why. That started in grad school with quantitative psychology. It took me to market research, where I’ve worked with large CPG companies, retailers, emerging small brands, survey research, data and analytics, and marketing analytics. I’ve worn all types of hats at this point.
Beverly: Survey-based data, understanding people. See, I want to get a PhD in sociology like Tom Davenport. I’d love to go back and get another PhD in sociology. It’s very interesting, isn’t it?
Geneva: As people, we’re interested in what we’re all doing and why.
Beverly: Yeah, I love it. Very nice. Now, where were you before? Have you always been in this field?
Geneva: My graduate studies were in quantitative psychology. Before that, I spent a little bit of time between undergrad and graduate school in rater training, project management, and the health sphere. A lot of my work was with studies of new treatments for Alzheimer’s disease on the project management side of things.
That gave me a lot of opportunity to also work with, in terms of the rater training, I was able to spend some time supporting research and aligning scores and performance and things like that. It was data adjacent, and that was part of what continued my interest in wanting to go into a quantitative field.
Beverly: Nice. I kind of started in the same place, like market research, data collection, survey research. I mean, my first model was with a utility company for cost of service. But once I got my master’s, I started doing a lot more surveys and understanding people and just seeing the data behind what they’re thinking. I always thought that was super intriguing. It sounds like that’s what makes you tick too.
Geneva: Absolutely. I like puzzles, getting in there and solving the mystery.
Beverly: So cool. Yes, I love it. So, when we think about AI, because everybody’s like, “AI, AI, AI,” everywhere. When we think about data science and AI, we typically think of a certain profile of companies. And if you go to conferences and if you hear people talking, the thought leaders, the books that are written, who are they talking about? Who are they actually talking about?
Geneva: You know, we’re usually talking about the big companies with huge data streams, and they’re thinking about scalability and populations, and we’re talking about samples.
We’re usually talking about the big guys, and there’s the rest of us who are mid-sized or smaller companies who are also navigating this landscape of AI and what we can do with it.
Beverly: This is huge, Geneva. Tell us if there is space in the small to medium business for AI?
Geneva: Yes, absolutely. I’ve been at my company now for under a year. Shapiro+Raj focuses on qualitative and quantitative, and the adoption of AI across the organization has been so impressive to me. There’s absolutely room for ways of working as well as analysis of qualitative and quantitative data. So, we’re on the frontier right now of some really powerful game changers for not just the big guys.
Beverly: Why now? Is there something special that has happened that makes AI maybe more approachable for the common guy? Or do you think we just, you know, kind of got tired of waiting in the wings? Why do you think now is a great time? Or is it just like, man, you better at this point?
Geneva: I think there’s a little bit of that. That’s kind of open to hearts and minds a bit too, you know, speaking that language where it feels like with gen AI and large language models and everything, that’s it’s a little bit more approachable. It feels like this is something that everyone can do. So, there is a bit of FOMO as well as just feeling like I can.
Beverly: Okay. So, I want to, and I’m a little jelly, and maybe I don’t have to be a PhD in computer science anymore. Maybe it’s possible to approach this from a different way, especially with the technology sort of being transformed to adjust to us instead of us adjusting to it. I think maybe that’s a paradigm shift, as far as how things are structured and how we solve problems.
When we think about small data or it’s not that they’re small problems, but we think about small datasets or smaller businesses or smaller types of things that aren’t teams that have teams, and giant enterprise-wide initiatives, and all these cloud solutions, and data science and AI all over the place. But if you think about the rest of us, can you give us some use cases where AI has been helpful?
Geneva: If we think about AI OG, and classic models like machine learning, we apply those with segmentation solutions, and you can apply machine learning methods for doing your classification or clustering and things like that.
But then, when we also think about new big AI tools, I think the biggest use case or at least the most accessible is kind of in ways of working. We now have this ability to talk to AI in a way that we talk to people, which makes it more approachable for a lot of the non-data scientists among us. But we can use it still to interrogate and aggregate across different data sources.
Whether people use it for things like going through a transcript of a 90-minute meeting, summarizing, and coming up with next steps, pain points, or in ways that we like to use it can also be looking at reports. So we work closely with clients, and we don’t have that spigot of data like big data and data science companies. But we have this trickle of reports and datasets that come in.
Beverly: But those are more precious to you, aren’t they? Like, doesn’t make it more precious because there are fewer.
Geneva: Yeah, they are babies.
Beverly: You have to get as much as you can out of them though.
Geneva: Yes. You really want to work with them and do everything that you can. So one of the things that we’re now able to do, where before you’d have to have people kind of pouring over them and taking their time and connecting the dots, you now have an AI research assistant that can take a look at all these reports. Not just monitoring brand equities scores across the years, but like, what changed and how does that relate to things that happened in the market or other elements of research? Let’s compare their brand equity scores and loyalty metrics. You can ask these questions now and bring these things together for you where it would have taken forever before.
Beverly: A lot longer. And to have to have a certain level of talent, cultural support, data availability, and more. So, it makes problem solving a little more accessible, if I’m hearing you right.
Geneva: Absolutely.
Beverly: Is there anything off the table that still as a small/medium, we can’t do this, or is sky the limit? Anything they can do, you can do better.
Geneva: Maybe a little. There are always standard differences, right? If you’re dealing with a bunch of responses that are the population of data, you’re not going to have sampling bias. So, I mean, there’s always concerns about smaller datasets.
Beverly: People don’t think about small data problems because big data problems, a lot of times, are volume, compute power, and that kind of thing. But in small data problems, you’re trying to get real value out of a small amount, and that’s tough.
Geneva: So you’re actually touching on one of the limitations with AI, which is, you want to be able to just keep asking them like, “well, what about this? What if you looked at that?” With limited features, essentially, not as many variables, not as many inputs, and limited observations, you get to the end.
Beverly: Oh, I see what you mean.
Geneva: There’s only so much you can get from the stone. So that is a limitation in a way and something to be mindful of. When you have limited data, there are limitations. But, outside of that, the ways of approaching it and thinking about it are still in development.
Beverly: Yeah, for sure. What do you see as some of the high value? Like, what kind of value can you get from this that was never really attainable before? Now that we can talk to AI and query things, and you don’t have to necessarily be a really strong programmer, does it produce more value for us? Or is it just doing small enough things that we’re still having to do the big picture?
Geneva: One of the nice things, but also maybe not a limitation and just a consideration with smaller data, is that you can provide more oversight to the model, so you can validate it a little bit more easily. You can go in and say, “you said this, let me take a quick look at that. Like, is that true?” It’s easy to get lost in that, and then you lose the value of using the AI in the first place. But it is also nice that you’re able to feel good about what you’re getting out of there, that it’s not a hallucination or a surprise.
Beverly: So, do you think it’s possible that AI can act almost as a wall? Right now, what we’re seeing is, with small data problems, people that work in this area, they already know. If you tell them this, and they say, “that doesn’t sound right, I can go look at the data and tell you.” They have an instinct about it. In a way, if you really think about big data, people don’t have that instinct, as well, maybe not as intimately.
Everybody should have a certain walking around knowledge about the descriptives of their data, especially senior-level people. So they would be able to say that it doesn’t pass their sanity check. With small and medium-sized businesses, they’re even more intimately connected to their data. Is it possible that AI is going to create a barrier or a wall where you can’t see that as much?
Let me give you a scenario. When I used to teach applied statistics at Berry College, I had my students manually calculate standard deviation and manually develop regression models. They hated me for that. But interestingly, they didn’t really understand what they were standardly deviating from.
So, they didn’t really understand why it was even called standard deviation or what it really meant until they did this manual process. Otherwise, they would just stick it in their phone or computer and go, poof, there’s standard deviation. Why would you make me do this by hand? And that’s why I made them do it by hand. Are we at risk of losing our sanity checks because of AI?
Geneva: My prediction for that would be that there will be a bit of that and then there will always be people who care about making sure that they’re actually in touch with the data, and I don’t think there will ever not be people running the analyses somewhere. I think that, at a certain point, when something new happens, you pull it back a little bit and kind of adjust.
Beverly: Oh, you think we’re going to go over the tip a little and then come back and be like, “wait a minute, maybe we should be programming.”
Geneva: That’s my guess.
Beverly: Does a small/medium business in some ways have an advantage to adopting AI compared to larger organizations?
Geneva: One of the things that I have been so impressed with is the level of adoption across the organization, and it’s because when it is adopted, it’s easier to make it part of the company culture.
I’ve seen in larger organizations you get this kind pocket of evangelists and maybe some people who think that’s cool that are around them, but that can just stay over there and other people don’t have to worry about it. When it’s a smaller company, you’re chatting with people who are like, “I just threw my transcript in this yesterday” and you know, it’s hard to stay away from and you can create more of a culture around it. So, I think there are some advantages as far as adoption in that way.
Beverly: Well, and also it seems like the buzz would be real because people could use it to hire the right person? It’s so tangible, and they can relate to it probably more so, because I’ve worked with large and small companies and everything in between, nonprofit, for-profit, all these different companies, and it’s funny to see the patterns that sometimes it seems like, gosh, if it weren’t for money, a lot of times the small/mediums have an advantage, I would say, over the enterprise big guys.
Geneva: In a lot of ways, there’s intimacy with the data, with the clients, and with the ways of doing things that I think is to the smaller companies’ advantage.
Beverly: Geneva, you’ve given us so much hope. What do you see as the future? Because so many people that are in small and medium are like, “oh, that’s not for me. I don’t get to do that because my company, they’re not going to do it.” And you’ve given us so much hope. What do you see for the future of AI for the rest of us? If you were to envision this world of companies of all sizes?
Geneva: So, what I would predict is that – the OG AI, like modeling things, you know, there’s some halo effect where people are more interested in those because of generative AI, and large language modeling, and all of that.
But when it comes to next-level AI, I think that you’re seeing it get integrated into not just data science and these things that us nerds over here are worried about. You’re seeing it in such a tangible way like how you work, how you interact with your day-to-day, that I feel like that’s going to keep happening as well as it’s just starting to happen in real life, like normal non-business needs.
You know, people are using AI to support things like writing a letter to the school board. It’s becoming integrated in these non-work ways, and it’s coalescing in a way. It’s coming from the work world and non-work world, and it’s all coming together like it’s going to be integrated. There’s going to be a point where it’s just what people are going to be doing. So, I think, small business or large business, we’re all going to be using some version.
Beverly: It’s funny because when I’m talking to people in a room and if you ask the senior leaders, what kind of AI and generative AI are you guys using? They’re like, “we’re not using it, and we’re being really careful and very slow to adapt.” Then they walk out of the room, and all the people that are director and below are like, “we’re using it left and right.”
They don’t realize all the different application areas. So there’s this notion because you keep talking about the kind of the OG, and I can totally feel you on this, because there is sort of a machine learning, and I said like a decade ago that we’re going to quit saying machine learning and it’s just going to be how we do data science. There’s not going to be this designation of manual modeling or whatever.
It’s just going to be how we do data science. But the sort of OG AI and the machine learning is kind of extending data science into AI, right? But you’re talking about the second path, which is an interesting idea because that’s a new level of user or developer. It’s the, as you said non-business or non-nerdy user, that are just able to use language and know how to query if you can ask the question in the right way.
Geneva: Yeah, like the engineers of querying. That’s a new expertise at this point.
Beverly: What final advice would you have? Because now we’re feeling all hopeful for all those people that said, “I’m never going to get to do fun stuff at my company because we’re not a giant enterprise. I don’t work for, you know, fill in the blank.” But what final piece of advice would you give to somebody trying to embrace AI?
Geneva: I would say not to put your company’s data in anything if they’re asking you not to, obviously. But you can start with those letters you’re writing and something creative that you’re doing. Like, just start using it somewhere. You want to fight your property tax increase. Ask Chat GPT or whatever to help you on how you might go about it or where you should look, and what the comps might look like. So just even personally.
Beverly: That’s funny because Mark Jackson, who’s a Tableau Zen master and works at Piedmont, talks about one of the best ways when he tells people they were really trying to understand business intelligence, this was years ago. He says to start with your favorite sports team. There are all kinds of data about your favorite sports team or whatever, but do something, so you’re saying, take it to the personal level or some other part of your life or whatever, and just at least start embracing it from that standpoint.
Geneva: Exactly. Start embracing it, and you’ll be able to speak to it a little bit more from a knowledgeable place when you want to advocate for it, so you’ll have that experience.
You can say, “maybe it wasn’t a business use case, but look at what I accomplished. We’re missing out on this.” If that’s something you want to be able to advocate for or just use. So maybe you’re not using it to evaluate your models or something, but you’re using it to help you write a report, where you’re not doing anything proprietary, you can still access it that way.
Beverly: So just try something, get comfortable, and then you can think about how to lift and shift.
Geneva: Exactly.
Beverly: Love it. Excellent. Geneva Taylor, thank you so much for talking to us about AI for the rest of us on TAG Data Talk.
Geneva: Thanks.
From small firms with limited data to large enterprises, AI is proving to be a transformative tool that can provide powerful insights and efficiencies. As AI continues to evolve, it’s clear that its potential to drive innovation, optimize decision-making, and enhance business performance is within reach for organizations across the spectrum.
Explore the full catalog of TAG Data Talk conversations here: TAG Data Talk with Dr. Beverly Wright – TAG Online.