Zach Hatch - The Future of Seamless Interactions with AI

The conversation this week is with Zach Hatch. Zach is an AI and automation engineer at Gemini, an advanced marketing solutions company in Edina, Minnesota. He specializes in building custom AI automation systems for businesses, marketing agencies, and anyone in need of optimizing through automated efficiencies. Additionally, he has real-world experience in building automated project management systems, social media accounts, note-taking transcribers, content generators, and much, much more. Before his current role, he was at Dream Home Furniture and Mattress, where he led and managed all aspects of marketing and technology.

If you are interested in learning about how AI is being applied across multiple industries, be sure to join us at a future AppliedAI Monthly meetup and help support us so we can make future Emerging Technologies North non-profit events!

• Emerging Technologies North
• AppliedAI Meetup

Resources and Topics Mentioned in this Episode

• Gemini: Advanced Marketing Solutions
• Zapier
• Visual Studio Code
• TLDR Newsletter
• Northlandtackle.com

Enjoy!

Your host,
Justin Grammens

Transcript

[00:00:00] Zach Hatch: I think we are still in that phase of being reliant on, you know, systems that we have utilized and gotten familiar with over the years. Large language models will make their way through integrations. For example, the Google landscape is actually currently changing with their generative search implementation.

I believe that we won't have to necessarily shy away from old methods of conducting research or finding information, but rather the old methods will adjust with the new technology. The large variety will see their interactions with AI be done more naturally because it'll just be. Seamlessly integrated into the services they're familiar with.

[00:00:40] AI Voice: Welcome to the conversations on Applied AI podcast, where Justin Grammens and the team at Emerging Technologies North talk with experts in the fields of artificial intelligence and deep learning. In each episode, we cut through the hype and dive into how these technologies are being applied to real world problems today.

We hope that you find this episode educational and applicable to your industry and connect with us to learn more about our organization at AppliedAI. mn. Welcome,

[00:01:11] Justin Grammens: everyone, to the Conversations on Applied AI Podcast. I'm your host, Justin Grammins. Our guest today is Zach Hatch. Zach is an AI and automation engineer at Gemini, an advanced marketing solutions company in Edina, Minnesota.

He specializes in building custom AI automation systems for businesses, marketing agencies, and anyone in need of optimizing through automated efficiencies. Additionally, he has real world experience in building automated project management systems, social media accounts, note taking transcribers, content generators, and much, much more.

Before his current role, he was at Dream Home Furniture and Mattress, where he led and managed all aspects of marketing and technology. Zach and I met a few months ago as we were both on a panel talking with small business owners about the various uses of artificial intelligence. And after having the chance to share the stage with him, I knew he would be an excellent guest on the podcast.

So thank you, Zach, so much for being on the program today.

[00:01:59] Zach Hatch: Yeah. Thank you, Justin. It's a pleasure to be here.

[00:02:01] Justin Grammens: Awesome. Well, good. So I gave a little bit of information on where you are today. I know we'll talk a little bit about Gemini and what you guys are doing around marketing, but maybe you could fill in some of the holes, I guess, with regards to how did you get to where you are today?

I mean, has this been a passion of yours? And maybe, you know, once you get out of school, maybe give us a little bit more context of your background.

[00:02:18] Zach Hatch: Well, it sure is an interesting journey on how I got here. My background was always in marketing. So initially I went to school for marketing. My time at dream home was spent, as you mentioned, that was kind of the head of marketing there, led everything in regards to.

You know, that facet, but through my journey, I essentially really had an interest in development. I would develop WordPress websites and, you know, modify, customize them, things along those lines. Nothing too in depth in regards to development. When I started at Gemini, I was actually brought in to assist in their paid media for the purpose of marketing.

How I got started in AI. Well, In the initial release to the public of ChatGPT in November 2022. I really kind of dove headfirst into generative AI, and it just became a really sparked interest of mine. From there, I think it was about mid to maybe end of January, that'd be 2023, so a few months after the public release.

I started playing around with the API. And when that came into place, I began automating social media accounts, uh, was how I really got started. So I automated a Twitter, a LinkedIn, an Instagram using Dolly, and a Facebook account. Now, granted, this was in the early days of generative AI, so it was riddled with hallucinations, full adrift, but It was more of a proof of concept than anything.

From there, I started kind of automating blogs and doing some WordPress automation type things and Gemini caught on. Gotcha. And there, the rest is history. I've been in this role for. Coming up on a year now.

Yeah, and so this is kind of a, I guess, a differentiator you guys feel within the service offerings within your company, right?

It definitely is, and you know, with generative AI, there are a lot of aspects of that that can coincide pretty seamlessly with the marketing landscape, so it really did make sense.

[00:04:20] Justin Grammens: Awesome. I thought it was interesting as you were talking about kind of getting into tech through WordPress, I mean, I think that is a great platform for people that maybe don't have to get into PHP code, but they can like use all these plugins.

It's very, very versatile in a lot of ways. I feel like maybe, correct me if I'm wrong, but the whole release of GPTs in general has allowed people just to sort of jump in with, they don't have to understand really a lot of the stuff under the covers. It'll actually, Yeah. It allows you to be super easy to tie stuff together, right?

[00:04:46] Zach Hatch: It really does, yeah. And I think anyone interested, you know, I always recommend just getting started. A lot of times AI can help you unpack the questions that need to be answered for you to move forward and become more successful in it. So that's kind of what helped me for sure.

[00:05:02] Justin Grammens: Yeah, it's just awesome that like these tools and large language models are allowing us all to level up in a lot of ways in areas maybe we're not so strong in.

So I almost feel like you don't have to be a code guru these days, right? You're able to sort of plug stuff together, either query, um, ChatGPT to get the code that you need to plug in. But in a lot of cases, we're making more and more like no code, low code ways for people to make these tools work.

[00:05:24] Zach Hatch: A hundred percent.

Yeah. I mean, Zapier has great interfaces where you can do a lot of no code solutions from there. In addition, if you do want to take the more technical route, Copilot baked right into Visual Studio Code can teach you more than I learned on my own in years. Yeah.

[00:05:41] Justin Grammens: What's an example, I guess, if a company were to come to you and basically say, Hey, I've, I just, I have a lot of manual processes that I'm doing.

Maybe you could just think off the cuff of like maybe one company that you've worked with, you know, just, it's all just generic stuff. I'm not asking for names or anything like that, but just, I'm always curious to ask people like either something that you're working on or maybe even something you've seen somebody else work on, like just the landscape is so broad and there's always new stories.

[00:06:03] Zach Hatch: Yeah, there definitely is, uh, one that kind of comes up pretty often and it's a, you know, a real simple solution is in the realm of marketing. A lot of organizations will kind of. Manually assess different CRM additions or different marketing leads. So a real simple one is automating that process and then essentially giving, you know, chat GPT or another generative AI, some eyes on it to be able to classify some of those leads, give you more insight that a human might not be able to, you know, just get off the cuff while looking at some of the.

Some of the leads that are coming in. So that's a real simple one that actually comes up fairly often. And you know, solutions like that can really bring a lot of efficiency and a lot more like insight into their leads are and their customers needs.

[00:06:48] Justin Grammens: Gotcha. So basically sort of taking a look at at a high level, these are the types of leads like the type of person that's sort of coming in.

Maybe great content and stuff that will help generate more of these types of people.

[00:07:00] Zach Hatch: One of the low hanging fruit ones that I would say would assist a lot of organizations who are having inbound needs coming in, or to go to kind of a larger scale, you know, in the instance where there are actions that are more tedious, more time consuming that not necessarily Really need to be consuming as much time as they do for certain individuals.

I'll give an example here quick. So you touched briefly on the process that I took to automate some of our project management operations, taking notes during meetings, that alone, following up after the meetings, things like that are really simple things that now AI can do for you. So we transcribe all of our zoom meetings, utilizing AI.

We use a separate transcriber that's powered by the whisper network. But Zoom even itself has a really good baked in AI companion that will also serve the purpose. But something as simple as that can really add a lot of time back to, you know, members throughout your organization and make you much more efficient.

[00:08:04] Justin Grammens: Yeah. So are you, you're starting to see, I mean, sometimes maybe you'll go into a place to consult with them and they might be able to just sort of use tools that are already built in, right. Rather than generating something new.

[00:08:13] Zach Hatch: Correct. Yeah. So I always advise everyone have, you know, in the world of AI, everyone's on the rat race to implement AI into their systems in some way.

So I always recommend, you know, regularly meet with your representatives for some of your platforms and systems. Chances are they have solutions that are already built in that you can start utilizing today and take advantage of, you know, generative AI now.

[00:08:36] Justin Grammens: Awesome. So some people are sort of scared of, of AI with regards to taking their job.

I'm wondering, Kind of what you're seeing and maybe what you see for the future of work and specifically maybe with marketing or any other areas you want to talk about, but what's, what's your sort of perspective on that?

[00:08:52] Zach Hatch: So my perspective is I don't think people have to be as concerned as they are essentially.

Yes, there will be a shift in the workforce. But it's going to be not much more different than shifts that have taken place in the past. I always like to use the example of, you know, like the assembly line, or when automotive started really relying heavily on automation to complete a lot of their factory work.

It'll be more along the lines of individuals working alongside of AI and augmenting their workflow versus AI taking over their job completely in its current state. And, you know, at least for the foreseeable future, AI will still need. Uh, a human reviewing the work that it completes. So maybe your role might just shift a little bit, but it won't go away.

[00:09:38] Justin Grammens: Yeah. I, I sort of share that sentiment as well. I think it's going to cause us to think a little bit differently. I mean, I, so I've been working in this space for a number of years, but it's interesting, I find myself kind of still gravitating back to old tools. Like I'll get a question and I'll be like, okay, I'm going to go out to Google.

I'm going to Google for that. And then I'll stop myself. And I actually have to kind of like retrain my muscle memory to be like, Actually, this is a little bit of a better question for a large language model or something like whether we chat GPT or Gemini or something like that. Do you still think we're in this early phase where like, you know, I guess maybe a lot of the population maybe doesn't even understand that there's a different alternative out here.

[00:10:14] Zach Hatch: They don't. I think we are still in that phase of being reliant on, you know, systems that we have utilized and gotten familiar with over the years. I think that large language models will make their way through integrations. For example, the Google landscape is actually currently changing with their generative search implementation.

So I believe that we won't have to necessarily shy away from old methods of conducting research or finding information, but rather the old methods will adjust with the new technology. It's more my thought process behind it. Personally, I use AI all the time in my everyday life. I mean, I have assistants right into my phone that I voice chat with to get information.

So, I think it depends on the user. I think the large variety will see their interactions with AI be done more naturally because it'll just be seamlessly integrated into the services they're familiar

[00:11:12] Justin Grammens: with. Did you see OpenAI's and their, their, kind of their 4. 0 version that just came out with their multimodal type stuff?

[00:11:18] Zach Hatch: I sure did. Yeah. I've been playing around with the Omni model a decent amount actually. And the biggest difference I've noticed is the latency. It's a lot, lot quicker. The generation seemed to be a little more robust. I'm a big fan of the multi modal approach, you know, one platform or one LLM that's capable of producing, you know, a variety of different mediums and accepting them as well.

I think it's great. Gemini did that not long ago from Google DeepMind. So it only made sense that OpenAI kind of caught up in a sense.

[00:11:47] Justin Grammens: Yeah, I just thought there was a, during their release video that they did, there's sort of like big thing I was watching it. And yeah, when you talk about just voice chat, right, they had this whole thing where they, they basically sort of talked to this AI and this assistant.

And it was very, very natural. That was the thing. I mean, there obviously were some little glitches along the way, the sound kind of dropped out. It kind of picked up other things, but I was super impressed with regards to just the conversational aspect. The voice sounding super real and it also just being, yeah, very, it flowed very, very well.

One of the things that they put in was that, you know, it doesn't have to sit there and wait for you to finish. It can kind of interject along the way, which felt very human in some ways.

[00:12:25] Zach Hatch: Yeah, I agree with that. And I really appreciate that they are trying to make it more human versus just robotics so that the term assistant doesn't necessarily fall on just, you know, a piece of technology sitting in your pocket, essentially.

[00:12:39] Justin Grammens: Yeah, yeah,

[00:12:39] Zach Hatch: for sure.

[00:12:40] Justin Grammens: Are you seeing marketers, I don't know, being ahead of the curve, behind the curve, kind of with the curve? A lot of people ask me, well, so how can I use, like I'm a small business owner, how can I use, you know, generative AI in my business? I'm like, well, if you haven't started using it just to write marketing speak, that's, that's like a first place to start.

But I still feel like there's a fair amount of companies that still are sort of hesitant. So I'm just kind of curious to, to me, it seems like low hanging fruit, a good place to start. But. Yeah. Wanted to get your perspective on that from a marketer.

[00:13:06] Zach Hatch: I would agree. It's very low hanging fruit. And I think for anyone, you know, emerging into the AI space, I think it's a great place to start.

It's about 50, 50, as far as adoption goes and integration goes, we. Are very in tuned with a lot of different agencies, you know, and individuals conducting marketing. And I'm not seeing as many adopted fully as I would have anticipated. Now we started adopting it rather early in comparison to, you know, a lot of organizations who essentially let a lot of the information kind of run its course before they started getting on board.

So in that regard, I would say it's probably about what we expect, but I would anticipate that Most agencies should be adopting it, honestly, I do think marketing agencies are familiar with AI in regards to it. You know, it's built into like Google ads or Facebook. So I feel as though a lot of organizations are waiting for those companies to take the step to AI instead of Exploring opportunity on their own, essentially, is kind of my thought behind it.

[00:14:16] Justin Grammens: Yeah, that makes a lot of sense. That's what I'm seeing with customers as well, too. It's like you're starting to see a little AI agent box be sort of put below the thing. And they're waiting for people just to sort of say, click, use AI to write this. You know, I also think there's something around prompting, right?

And that I think people don't understand the power of the prompt. And being able to tell a story in your prompt to set the model up with what you want to get because if you can give a good story sort of behind with the output that you want, then you're going to get better results. And I, I feel like some people just expect it to magically know all this stuff, but you sort of say, generate me copy for a blog about X, but what is the tone that you want to have?

What are the specific subject matters? What is your audience? Those are types of things that I think will help. So you got to kind of understand and you find yourself maybe educating people on some of those aspects. Yeah.

[00:15:03] Zach Hatch: Yeah, I completely agree. I don't think people understand the power of a prompt and prompt engineering is in a way, almost a science, you know, understanding how to formulate the fuel for the AI to get what you want out of it.

I think that it has more power than people realize. Now, starting out with AI, obviously your prompts won't quite be there, but it all comes with experience. There are certain frameworks individuals can use that I was kind of like to lean on and tell people to, you know, this is a really good starting point, which always helps with generations as well too.

So I would agree there. And I think that, you know, we're still in a state where prompting is very important and in the future, you know, maybe how you engineer a prompt won't necessarily be as pertinent, but for now it's, it's the fuel to the fire in a sense.

[00:15:52] Justin Grammens: Yeah, that's a great point. I just was just having a debate a couple days ago, actually, I think it was at our conference that we had last week, but people were, I think it's 50 50, it's kind of mixed, like the system is getting smarter and smarter.

So at some point, the prompts, maybe the value of them are going to become less and less, right? That's kind of what people are saying over time. And so this whole idea that you're gonna need to sort of cater to the AI in some ways, hopefully the AI can start to cater to you, but difficult to know what's going to happen, sort of play out at the end of the day.

It is difficult. So we will have liner notes, you know, with regards to our conversation here. So if there's any links afterwards, Zach, you want to send me or whatever, I will definitely post those. In the podcast notes, uh, one of the things I do like to ask people is, is, you know, imagine yourself maybe leaving college or whatever, just getting into the field, like what are some things that you would suggest people do?

Are there books that maybe you could read or good websites? And again, I'll get a list of all that stuff, but anything off the top of your head, I guess you'd be like, Hey, this is how I got into it. Maybe this is how you should as well.

[00:16:46] Zach Hatch: Yeah. Well, I wouldn't recommend how I got into it. It was a lot of struggle, but for someone new getting into it, the number one thing I recommend is TLDR.

They have incredible newsletters that they send you every single day. You can sign up for a variety of different verticals within their newsletters, that being AI, development, marketing, you name it, they have something across the entire gamut. By just signing up to that, that will give you An incredible headstart in front of anyone.

You'll be fully up to date with the news AI. They give you tips on prompt engineering. They give you tips on more advanced things. If you are going down the development route, you can go towards some information surrounding reg or how to make your calls, things along those lines. So that's a great place to start from there, realistically, get your hands on it.

Um, start playing with the technology, essentially, see what you get out of it. And hopefully that'll just kind of get the wheels turning towards what the possibilities are and how it can be implemented in your organization.

[00:17:47] Justin Grammens: Yeah, great, great, great. Very cool. How do people get ahold of you? What's, what's the best place for them to contact you?

[00:17:52] Zach Hatch: Well, LinkedIn's a really great place. So connect with me on LinkedIn. I'm always happy to, you know, grow the network. We also post some AI news updates on our Gemini LinkedIn page, so you can follow us there and. You'll get to see me pretty regularly if you're following us there as well too. So I would say people can connect with me there.

[00:18:10] Justin Grammens: That's great. Yeah. Do you have any, uh, you know, you and I met at a speaking thing with, uh, Chamber of Commerce. Do you have any other events that you're speaking at

[00:18:17] Zach Hatch: coming up or anything? Nothing coming up soon, uh, but I do have some things on the books for about a month and a half out. A lot of the speaking events that I currently do aren't necessarily public events that people could really join.

But when I do, I'll be sure to let the network know. Well,

[00:18:36] Justin Grammens: yeah, we'll, we'll put a link to your LinkedIn page. I guess people can connect with you there for sure. Are there any other aspects around what you do that maybe I didn't, I didn't touch on that you would want listeners to, to know more about? One big

[00:18:49] Zach Hatch: thing that I guess we didn't touch on is my Big specialty currently is actually building chatbots.

Automations are my passion, but AI chatbots is something that we really started kind of implementing for various customers and things along those lines. A resource like that that's trained on some of your information and your data can prove to be super valuable for both customers and internally for Maybe customer support or something like that to get questions answered.

So I would say that's just something that I kind of want to be on people's radar because there are small implementations that can be made that can really make a tremendous difference. We have some cool projects in the works right now. If anyone wants to check it out, we have an operating chat bot at northlandtackle.

com. That's one of the big projects we're working on. Uh, it's uh, you're a localized phishing expert, any phishing questions can be asked there.

[00:19:38] Justin Grammens: Awesome. Did you guys kind of build your own there or are you leveraging some third party off the shelf type things? We constructed our own. Gotcha. I think that's very, very valuable.

What's interesting is I think, you know, a lot of people have seen chatbots on websites for years. I mean, Google has had their, uh, dialogue flow or something like that. I remember experimenting around with that for quite some time years ago. But this is totally new, right? I mean, with regards to large language models and just the very, very natural feel, right?

What's sort of new here that you're telling customers that they should reinvest, I guess, in a potentially a chatbot?

[00:20:09] Zach Hatch: Yeah, no, of course. So, the biggest difference is old chatbots were based on a conversational flow that was predetermined. Based on, you know, buttons you might click or what you anticipate the chatter might be visiting for or require.

Now with the implementation and incorporation of AI, that process has completely changed. Now it's a full conversational flow that is, you know, natural to human language. We train them differently. So. We're focused on creating industry experts versus customer service representatives. So now we can create a landscape where customers can obtain valuable information about the industry you're in at an extremely high level, right from the source.

So essentially taking years of your knowledge that you curated and painstakingly earned and putting it into an AI chatbot that your customers and associates essentially have access to 24 7. We're finding that it's assisting with essentially converting more lifetime customers, increasing their value and really kind of just helping the organization in ways we didn't see before.

[00:21:16] Justin Grammens: Yeah, great. Awesome. Well, that's the whole purpose of these tools, right? Is to have them work for us, as to take out all the mundane stuff. And if we have a sales assistant agent that's out there 24 seven, that actually knows our product can actually speak to it as good as any sales agent could for us.

Then that's a huge win, right? It is. Excellent. Well, cool, Zach. I appreciate the time today. Thank you so much for being on the program. I'm super glad that we met and yeah, I look forward to seeing you at future applied AI events and other things we have going on, but. Keep in touch and thanks for sharing all of your

[00:21:48] Zach Hatch: knowledge with our community.

Of course. Yeah. And thank you so much for having me, Justin. It was a pleasure being on here and I look forward to connecting again in the future soon.

[00:21:55] Justin Grammens: All right. Take care. Thanks.

[00:21:58] AI Voice: You've listened to another episode of the Conversations on Applied AI podcast. We hope you are eager to learn more about applying artificial intelligence and deep learning within your organization.

You can visit us at AppliedAI. mn to keep up to date on our events and connect with our amazing community. Please don't hesitate to reach out to Justin at AppliedAI. mn if you are interested in participating in a future episode. Thank you for listening.