AI in Action: An Expert Roundtable

This webinar is part of Promevo's four-part webinar series on AI in the workplace. Use the list below to browse the other sessions. 

• The Executive's Guide to AI in Business

• Embedding Al in Your Product: A Promevo & Google Collaboration
• Ensuring Al Adoption & Regulation in Your Organization

Transcript

Brandon Carter:

Welcome in. We've got a decent audience here. So, good to see everybody.

I'm Brandon. I am the Marketing Director at Promevo. You guys aren't here to hear from me though. You're here to hear from these experts, all of which have done various incredible things with AI.

So I'm going to have them introduced themselves, starting with you, John.

John Pettit:

Hi, John Pettit, CTO here at Promevo. Help a lot of our clients and customers out there either onboard AI or build solutions for them. Super happy to be part of a stacked panel this week.

Brandon Carter:

For sure. Those of you that have viewed our other webinars in this series, John's been on every one of them and always has some really interesting comments and a lot of perspective.

Haritha.

Haritha Yanam:

Yes. Hey, everyone. Thanks for joining us today.

I'm Haritha Yanam. I'm founder of a Boston based AI startup called Ace4 AI. We specialize in building industry specific AI powered document intelligence solutions and I'm really excited to be here. Thank you.

Brandon Carter:

For sure. And excited to hear more about Ace4.AI as we talk through this roundtable.

Jump in there, Nic.

Nic Bryson:

Yeah. Hi, I'm Nic Bryson, founder of Orgnized.

After leading teams across the world and kind of the B2B SaaS industry for the last few years, realized that managers need a better tool to keep track of their one on ones with their team members. And, with the AI technologies today there's a chance for them to, really augment their efficiency and productivity as they manage larger teams and are asked to do more and more as organizations are leaner today.

So excited to be here as well. Thanks, Brandon.

Brandon Carter:

Love it. And Orgnized, very cool tool. I recommend checking it out.

And Trevor.

Trevor Turnbull:

Yeah. Thank you. I'm a co founder of a company called SinghBull Ventures, which I co-created with my wife actually. And what we do is we operate as Chief Fractional Growth Officers for the visionary founders that we work with who have big ideas and they don't know exactly how to pull it all together, whether it's the sales or the marketing or the operations or a little bit of everything.

So we come in and they bring us their ideas and we say, we'll go figure that out. That's obviously how we got introduced in working together as well.

Brandon Carter:

That's right. That's right.

And Trevor has helped several companies, whether it's from an employee perspective or from a product perspective onboard and build AI applications.

So basically with our panel here, you've got a lot of different approaches. You've got John, who as part of Promevo has helped on both the employee enablement side, as well as things like Vertex and helping companies build AI into their products.

You've got Haritha and Nic who are in essentially, like, startup environments where they're building AI centric tools and bringing AI centric services to the marketplace.

And then Trevor kind of does a little bit of everything helps build and helps companies adopt. Really unique perspectives and I'm excited to hear from every one of you.

But first a quick pitch about Promevo, like I mentioned, we are a full Google shop that helps with all things, Google Cloud. So that's your Google Workspace. That's your Chrome devices. We have gPanel, which is our proprietary Google Workspace user management and reporting tool.

But specific to this thing, we're helping a lot of companies with Gemini right now. Companies know they need to make an investment in an AI tool. If you've got Google Workspace, Gemini slides in extremely nicely.

But, there is complexity there, so we can not just help you purchase it, but help you get it engaged and get people using it throughout your organization. So reach out to us for all your Google Cloud needs.

So those of you who have been following along here, this is the final, the fourth and final in a series. We've been talking about all things AI for three or four months now, and AI is, it's not just an employee tool, it's not just. functions that people are adding into their apps. It's really transformational across a variety of, regardless of what angle you're looking at it. All of these are on demand and free.

If you want to go back and watch our first one where we're talking about AI for employees; the second one, we spend a lot of time in AI within a product environment in either like building things yourself or leveraging something like Vertex AI. And then last month we talked about AI adoption and regulation.

Companies are spending a ton of money to bring on these tools. How do you get your money's worth out of them? How do you make sure that they're improving your efficiency and, making your employees lives better.

So all of those are available. If you go to promevo.com/webinars, you can view all of those for free.

With that, let's remove our slide. Let's go ahead and start talking about AI stuff here.

So, two years ago, we weren't really talking a lot about AI. I know it existed. I know it was a thing.

And now we're talking to a panel where several of you have founded startups. Several of you have revamped your careers around it. How did this and John, I'll start with you. Like, how did this come about? Like, where did this sudden onslaught of AI interests, where did that come from?

What do you, what in your opinion...

John Pettit:

I think like all things, like it takes a while for water to boil, right? So AI has been under development for decades and. Transformer models that Google put out in the late 2017, 2018 really started transforming the power that we could get out of a neural network just predicting words and how deep that could go.

And with compute power and just continued evolution of data, we've now reached this point where. You can get real meaning out of a large language model and use it for a lot of different problems. People probably using it for things they shouldn't, but you can use it for a lot of different problems. And, it's starting to add a more simple to approach way to solve programs that aren't discrete, that have probabilistic answers that you can have many possible outcomes, and this is just one of those things, the frog in the water didn't know it was boiling until it's boiled. And now everybody knows they need to jump on this and do something.

Haritha Yanam:

Yeah. Just adding on to what John said, right?

So LLMs are not new. We have been working with LLMs and transformer models for a while now, but I think the big boom is with the generative AI models that it can be programmed in natural language, right? We can talk to the models in natural language in a conversational pattern rather than complex code.

So this ease of use and speed of outputs, I believe, led to mass adoption and the big boom that we are talking about now.

Brandon Carter:

So it was simplified in a way such as the average person, someone like me can go in and interact and build things and do stuff with it.

Would you agree with that, Trevor? Is that really what's happened? And now it's this gold rush.

Trevor Turnbull:

Yeah, for sure. I would say that, on the boots on the ground aspect of this, I work with lots of different clients that are always looking for the latest tool, and the hack, and the strategy. But at the end of the day, they want increased revenues, improved efficiencies, and savings in their business.

So I think what I saw in the last couple of years was just the language of how AI could support that started to become more and more believable, I guess you would even say, right? That previously it might have just been too technical and people weren't paying attention to it. But it's accelerated really fast though. I gotta say that.

Brandon Carter:

The way that I think about it in my head is like early two thousands, we had this mass adoption of SaaS, and everything was all of a sudden in the browser. And then over time it's taken about , 20 years, but everything like there's become market leaders, there's become standards, and it's condensed down into, a handful of tools. There's still a ton of innovation and new things happening. This feels like the beginning of another one of those sort of like rushes.

Okay, anyone can build SAS now. Now anyone can build AI. And you have people like Nic who. I have an idea of something I can do with this.

Nic, talk about like, how did you, like, how did you decide to build Orgnized or what was that tipping point for you? It's, I think I've got something unique with AI here.

Nic Bryson:

Yeah. It's funny going back to the previous question. I'm also thinking about let's say three or four years ago, we were all talking about the blockchain, right? This webinar series would have been about blockchain.

I think. It's one of those things where ChatGPT was the thing that really the fire where it became accessible for a normal, non technical user And I think that's where the real tipping point was in the last two years of why we're talking about AI now.

Obviously, all of the implications of the chip technology that Nvidia has popularized and but you know as we're looking ahead it's interesting because SaaS is now just a commodity. A fully featured SaaS application is table stakes. And so it's about what can you do on top of that, that has really a lot of potential.

And what's interesting is I think we're still in the place today where users don't necessarily know what they what they could achieve. And it's a really interesting time where a lot of innovation is happening. People aren't necessarily asking for AI features yet. And and I think that's where the, the really interesting applications are that we're exploring with Orgnized.

Brandon Carter:

I'm, that's a really interesting comment that you make there because I was thinking about it.

Almost every tool that I use now has an AI function. Is that right? Do you think that this is just going to become the norm?

There are stats out there and I don't have them available. I apologize. But Trevor, is it just going to become standard that you purchase something and it's going to have kind of an independent thinking aspect to it, whether that's like grammar fixes or analysis, is this the new norm?

Trevor Turnbull:

For sure. And we're looking for it in tools that we're using as well. And also, I think being cautious about making sure that the information that's being delivered is actually accurate, and in my case, when we're talking about creating marketing materials or improving sales process, just that it's in alignment with the brands and the companies that we're working with so that it doesn't go off the rails.

There's a lot of work that's put into refining those messages and AI can be incredibly supportive, but also take you off on the wrong track too.

So hence the prompting. This is why it's so interesting to be on a call with people with a little more technical knowledge than me too, because I'm always learning.

John Pettit:

I think it's interesting, Brandon, what you're talking about is like evolution of user experience. Nobody really asked for a web based forms on data thing to solve their problem.

Like they were like, I need to be organized or I need to generate reports or I need some sort of output. And so SaaS just created every industry specific forms on data solution you can imagine, and tried to iterate productivity through user experience design as far as they can.

But really that's, that the next gain of productivity is going to come by saying I don't have to spend as much time in that user experience, right? Like it's easier if I can just ask a system "tell me what happened in the data for the last 12 months." And it gives me back a chart or table, or, just update this record with some information instead of me having to navigate through 10 clicks to get to the page that I have to update something.

So the more AI starts to take hold in these applications, I think the more it's going to transform just how we interact, I think, Haritha said like in natural language. But all the systems that we deal with every day,

Brandon Carter:

Is it fair to say that, today we're seeing everything imbued with AI, but over time, theoretically, some things are probably going to go away.

One of the things that you mentioned right there is being able to take like a large data set and do analysis on it. And there are tools that do that, maybe not with AI, but now that may just be like a baked in part of Gemini for Google Sheets or Gemini for Workspace.

It's... theoretically, we're going to see some industries or at least some applications wiped out. Is that, do you think that's fair?

John Pettit:

Yeah. Why data replaced? I think it's just a matter of people are going to start thinking of what's the AI suite or platform that they need to buy and less about the productivity SaaS application that they buy. I think it's just going to be a pivot in mindset that happens over the next five years, especially as.

Companies are springing up like Orgnized or like Haritha's company that are providing a lot more capabilities around the AI frameworks that exist that people can then leverage.

Brandon Carter:

Another thing that I think Trevor touched on a little bit. AI these days, it seems like it's about 80 percent accurate. Like it's known to hallucinate a little bit and make some stuff up.

I'm curious, especially Haritha, who's working out of like legal documents is a major part of Ace4. Does that give you pause? Do you think about that? And what are your thoughts on it?

Haritha Yanam:

Yeah, definitely. One of the reasons why we got into legal document intelligence was there are only a few tools available in the market.

And there, there is always this question, right? The AI is 80 percent accurate. It's not completely accurate all the time, but I think one of the things that I really uh, like, about some of the practices that we are doing is the adding these responsible AI practices, right?

For example let's say some legal professional is chatting with the document, right? So our tool actually takes documents, legal documents, and then you can chat with your documents similar to ChatGPT but the difference here is we have so many industry specific guardrails that we add. For example let's say, who is the defendant? Who is the accused? Who is the judge? The answers are coming along with the context from which the text was taken, right?

So basically trying to make sure that we follow the responsible AI practices and then explaining why a decision was made. I think that's something not just for legal for any industry that we are taking. We need to make sure that the responsible AI practices are in built and governance and strategies are in bill so that we get that trust from the stakeholders. I believe I think that will help us.

But AI is not going to be perfect. That's something we already know. I think adding these governance strategies will help.

Brandon Carter:

Nic, you, your AI is a little bit different at Orgnized AI. Whereas what Haritha is doing is based on like the document itself is the language model. Yours is, I know one of the main functions is providing like management tips based on like inputs that a manager may put in like one to one meeting notes.

Do you like, do you think there's a risk of hey, this thing, it might say something crazy. One day, like I, you must fire everyone. I don't know. I'm just making stuff up.

But is that like possible or do you feel confident that no, the AI, we have it trained in to do what we need it to do.

Nic Bryson:

Yeah. We're, actively training it to ensure that it does give the right kind of outputs. And I think that's the key, right? So in, in your base LLM, AI model, there's no accounting for what's the gravity of being wrong, right? And it will take all of the inputs if you don't train the model down to something specific.

From a user standpoint, having spent a lot of time thinking about user adoption in my career in different roles. One of the things that is interesting to think about when you think about adoption of AI is people don't trust a black box. And so I think that is one of the kind of fundamental things that the LLMs and the AI tools that are out today have to overcome is that people have to understand that what they're getting is trustworthy.

And I think that's where a lot of the tools are playing today, right? When you lose, when you use a lot of the tools, you see an article about a new AI tool that's come out. You go, you sign up for it. And it's honestly, it's pretty shallow in, in some of, what it's capable of or different things.

And, the tools today are just emerging. Everybody's still experimenting, but it's all about the proprietary training of those models. The different ways... there's the title now of just "prompt engineer," when it comes to an engineering organization and how they're prompting the AI to get the right outputs and different things.

I think about it though, like I almost, I was ahead of this webinar. I was thinking about, the transition from the Manila envelope that gets passed around the office to email. And this idea of just massive productivity gains.

And I think that's the threshold we're on today, with kind of the AI piece it's a step change in what we're going to be able to achieve.

Haritha Yanam:

Yeah. And to add on to what Nic has said, so I think adding these guardrails. For any LLM products that you're trying to build, right? For example, you take output from an LLM, but you just not, you're not going to give that whatever the output the LLM gives directly to the user.

Rather, you have your own guardrails, right? You have other proprietary models That are going to verify the answers and then you have some ground truth data that you build or guidelines that you build and then provide the answer to the user along with the context, right? I think adding those car rails really helps.

But again, we are not there yet.

John Pettit:

Yeah, I like those concepts that you guys brought up around grounding. Like people have gotten used to ChatGPT and Gemini, just giving them answers and they don't realize behind the scenes. It's not just the LLM. There's grounding services. There's additional functions and tools running.

There's like a whole ecosystem of tools that they're hiding that you don't really see, but you get the answer. Even if it's go out to this webpage and scrape some data. That's not part of the LLM. It's, it's creating a sense of live information flow into the LLM.

And I think one other point is like, there's this set of industry data that you have available to you that people are now bringing into a database to make available to the LLM and its own, it's been its own way to understand it.

So, how do you organize data, get it semantically structured in a way that you can inject it into an LLM and create, a memory store, right? Like the human brain is yes, we can predict and anyway, we have our same neural pathways that we can, say the words that we're going to say, but, we have to inject context and memory into those.

And so this evolution of LLMs on top of contact systems like RAG databases or vector databases, or those tools that can be injected in terms of real time APIs, I think is something that people don't see, but it's extremely powerful in terms of how we're creating more accurate LLMs.

Haritha Yanam:

Yeah, completely agree.

The RAG systems and then doing the semantic matches on the questions that you're asking, they're pretty powerful. And that's something we, I think, we use in our products as well, which I think makes more sense to give the answers in more at least the closest accurate answers possible for the user.

And on top of that, I think adding the feedback loop, right? So you should always have the SME review the output, similar to what with ChatGPT, right? The thumbs up, thumbs down.

So you are giving the feedback, and you are iterating over the model to be retrained or in your solution to be retrained, really helps as well.

John Pettit:

Yeah, Nic, you said measuring their critical of the decision that human in the loop is important. So like you're doing like a medical LLM, you probably want to make sure you're using like a medical distilled model that is very specific to that case and still have the ability to have some people review accuracy, especially if it's giving advice.

So yeah, criticality decision matters a whole lot.

Nic Bryson:

Yeah, I think the sensationalism of the news cycle will quickly jump to our jobs going away or different things. But, it's the LLMs they're not critical thinking, right? They absolutely need us to be guiding them in order to produce the outputs that we're looking to gain for our efficiencies. But, it's not critical thinking.

Brandon Carter:

So I think this is a good transition because what I'm hearing from everyone is this is not like a plug and play sort of function. You don't just, your Lord of the Rings meme, "one does not simply add AI" to a product or to, a service it needs to be thoughtfully it, it needs to be thoughtfully procured and built and then iterated upon.

So I think with that, I would love to hear. Trevor, you mentioned a project yesterday. As much as you can share, I would love to hear about some of the things that you all have built and like, how did you account for some of this?

Trevor Turnbull:

Yeah, for sure. I don't claim to be overly technical when it comes to all of this stuff. I can, sit in the room and understand the language a little bit, but I don't build this stuff.

What I do is I bridge the gap between what does, what do we want to solve and how can the technology do that? And that requires a lot of training of people and vetting and making sure that the information is accurate.

The project that I worked on and worked with you guys actually on was to improve the efficiency of creating customized proposals for potential clients.

We had a lot of different data sources. So we had forms and Google Docs and then recordings from live calls by Zoom connecting to other tools, and we wanted to pull all that information into one document and concisely write out a customized strategy so that it resonated with that person, and you could tell that there was a layer of care that was put into putting that strategy together. So it wasn't just an off the shelf kind of vanilla template.

And what we found was that we could get that 80 percent accuracy and the 20 percent we still had to massage, and where this had practical application was, for example, if the person was filling out a form and also indicating who their wife was, or their husband, who would be involved in the decision making, that person was mentioned in the strategy documents, and if that name was wrong you've instantly lost credibility with that person, and that's just a micro thing, but it can be a huge thing in a long sales process.

So from the point of them understanding that there is an opportunity for them to increase the revenue or create new revenue sources for people, working in the corporate world, that want to own a business, for example, we needed to be able to communicate that in a way that allows them to educate themselves over time.

So it's a very detailed sales process. It takes months and months. And that document that we co created together has 10x the efficiencies of the guys that are sitting in the seats of as the consultative salesperson. So what that looks like is what used to take them 10 hours now takes them an hour.

And the information is way more rich and personalized, which is increasing the quote unquote close rates on the amount of deals that they can do.

So there's huge upsides to it. And at the same time, there is a layer of human interaction that's just absolutely necessary still. Otherwise it can go the opposite direction, which is obviously what we don't want.

Brandon Carter:

That's fantastic. And I love that measurable in there where, like, it's 10 X, you used to have to do this thing that drove you nuts and now we've built it out.

I, I am curious, like at what point did you make the decision, all right, I could use help in here and you tagged into Promevo not to make this any sort of sales pitch or anything, at some point you thought I could use a little extra help, like where, can you talk just a little bit about that?

Trevor Turnbull:

Yeah, I'm like your typical user that we've been talking about. I explored ChatGPT when it got into the news and because I'm more of an early adopter than the people that I work with, typically, I was able to play around with some stuff and go, whoa, this is cool.

But there was still missing pieces because we had all of these different data points coming from different sources and it wasn't all text based, right?

I couldn't just prompt something with text. We had audio that needed to be transcribed and then put into a machine and then just pulling it all together. It was just above where my technical knowledge is.

It was my wife, actually, that started doing the exploring of how we could pull all these things together. And that's how we came across you guys was through through Google Gemini, I believe, I think they had referred to you guys as a preferred partner.

And yeah I know we're not here to give you guys a heavy pitch, but the experience was great. The guys that came in. John, I believe you were on the first call that we did.

And then the guys that we worked with, they just help break things down into the language that we could understand. But then also, at the end of the day, what we wanted was just a working product. And that's what we got.

And I fully understand too, that things are likely going to break and change and need to be updated over time. But it's just exciting to see what can happen when you get the right people in the room, trying to solve the problem together.

John Pettit:

And I think Trevor, one of the things that made your project successful versus some of the other ones that, that you hear about out there is that you guys had a very clear KPI that you wanted to achieve.

Like you knew at the beginning. This process takes us 10 hours or something, we need to make this faster. Like you were very clear on what the goal was. And so that made it easy for our team to come in and give suggestions and even help direct things where maybe we're like, maybe you could do this differently here or there.

And same on our side, the team thought it was a great experience working with you.

Brandon Carter:

Fantastic takeaway. Big bullet point right there. Know what problem you want to solve and know a way to identify has this truly solved this problem?

Can we, can you talk a little bit about Ace4 and just as much as you can share about hey, this is how I built this thing?

Haritha Yanam:

Sure. Yeah. Ace4 is four month old company. So basically the idea was I worked with several companies where I see that document extractions and document intelligence is the key for many of the companies. I worked with John as well earlier. So there is a lot of document intelligence requirements across the companies.

But I think one of the industries that I'm really interested in was legal where I didn't see too many products for document intelligence. And e-learning is one other industry which I'm really tied to where I'm also an adjunct professor part time at UMBC. So I teach quite a lot.

So I think I know some of the pain points where I need to sit for hours to create the quizzes or create the flashcards, and going through the same material and revising, if I'm going to a different subject, right? So I think, those, some of those those things tasks that I'm doing, I was doing, I wanted to automate this.

I tried experimenting this with some of my students I was working with that UMBC, and we tested this out and definitely helped in saving time and improving the efficiency for e-learning. So that's when I decided that, okay, this is something I'm going to build at scale.

So now with the product that we are releasing, anyone in the e-learning industry are able to take any type of document, be it audio, video, or your PDFs or PPTs, whatever the document formats are, they are able to take them, convert them to text and do different tasks, right?

I want a lesson summary. I want a flashcard for my speech. Specific lesson, right? So things like that, you can automate and get extra insights from the documents. So similar to that, for legal we are doing like, brief generation evidence gathering and a lot of different tasks, which lawyers spend a lot of time reviewing those documents.

And at scale, right? So it's not just about document intelligence. It's about managing those documents together, right? If you have a team of lawyers using this tool, so you are, our admin can manage those documents for you. So everything is in place along with the raw documents and outputs.

And then you have the option to use the tool on demand, right? If you have a hundred documents per month, you can use the tool for 100 documents so you can scale as you grow.

So I think I'm really excited about this tool. I strongly believe that it will add a lot of efficiency gains for for the industries.

Brandon Carter:

Could you, without exposing anything proprietary, can you talk a little bit about what's the stack behind it? What are some of the different tools that you've had to interact with or build to bring it to fruition?

Haritha Yanam:

Sure. Yeah. There are some so for data conversions from audio, video, and PDF. PDF is a standard libraries that we can use, but for audio and video, we have proprietary models that we train on for legal separately and for the e-learning separately. So those are proprietary trained models that we're using on top of the LLMs, right? So we fine tune their alarms that that are existing.

And then there are some, it's not, everything is not about AI, right? There are some extractions or insights that you can pull with regular expressions or simple tech stack rather than going into detailed, complex models.

And then we have fine tuned legal LLM that proprietary LLM that we trained. Using some of the, legal documents. So we are using that to do some of the processes. So I would say that tech stack is proud. Obviously, we are, we experiment with Google cloud and AWS as well. And I think one thing that I really like about Vertex AI, right?

So I think to test out all these concepts at the beginning and to see if this actually works or not, we experimented with Vertex AI to see, how is it even feasible? So I think that really helps if anybody wants to get into the AI product development, to just test out your features and see if the MVP will be out the door.

I think vertex AI or SageMaker, whatever tools you are you are comfortable with so you can use those. Tech stacks are like proprietary models, some LLMs and some software engineering and a lot of software engineering practices. Yeah.

Brandon Carter:

Just as a, for those of you watching now and in the future, go back to number two in our AI webinar series, we really go in depth on Vertex and some of our developers are on there talking about like how it can really give you a leg up in building AI functions.

And yeah, really neat stuff. I think, what's interesting, Haritha there's a lot of excitement around, like language editors, doc the things that you see online " help me write this document," "generate an image."

But I think really some of the most exciting opportunity in AI is taking industries that we felt were " they are what they are" and imbuing them with something that's very different, very new, like legal documents, how much how much excitement is there? It's it's been what it has been for forever, right? Like it just stinks.

But this is where there is massive room for disruption. And AI, when you talk about like gaining efficiency, the whole, pick any part of the legal process, like any of that is just ripe for it. Massive opportunity.

Haritha Yanam:

Yes. Yes. And one other thing I forgot to mention is none of the documents that users are going to upload are going to be used to retrain any of the models. I think that's a key... that's what users want, right? They don't want their appropriated data to be used to retraining the models. I think that's something, yeah, we are. So a lot of responsible AI practices are involved to build the tool.

Considering legal, right? So we want to make sure we're careful.

Brandon Carter:

Yeah, absolutely. And we'll talk a little bit about security again here in a minute. But that's a big one. That is something to be concerned about.

But yeah, thank you for walking through that. That's fantastic.

Nic, I would love to hear about Orgnized and your journey.

Nic Bryson:

Yeah, you know, it's been interesting as we've been building our prototype. And in all of the user interviews and research we've done, 85 percent of managers are using Google Docs or Evernote to keep track of where they are at with their work with what we talked about last week, "oh, I need to talk to Brandon about this tomorrow."

And, it's interesting at, to think about the potential efficiency gains from going from, uh, basically dumb note taking tool to piecing that all together around where a manager intersects their day to day.

And in all the work we've done to understand what product should fit in that gap. No one is asking for AI features.

And it's interesting. I think what Trevor mentioned was super important, which is being an early adopter, testing things out is the thing that's going to put some people above the rest when it comes to understanding how to use and adopting these newer technologies today is the time to be testing those things out.

And so as we've, as, and as we've developed, we've been looking at very specific use cases where, people don't need, again, like a black box where they don't know what's going on behind the scenes, or they don't feel like they're in control, but rather, Hey, I need these notes summarized for me, or I need that reminder three minutes before the meeting happens with two paragraphs of information so that I can walk in prepared and then I need it to transcribe it for me when I'm in the meeting and follow up with me with the tasks in my work management tool after I leave.

And all those small use specific use cases are where, today, that's where the AI implementations can really thrive in those very distinct ways as opposed to, here's a broad tool, play around with it.

I need to get my work done, right? I have a, I have a, a need and that's what I'm here to do. I'm not here to use AI for the sake of using AI. And I think that is really where, building the right things comes into the user journey.

Brandon Carter:

Yeah, again, a big bullet point.

John Pettit:

When you guys think about that, Trevor, you and Nic both mentioned it is like this human early adopter and retraining, right?

Do you think about that when you think about who the people are you want in your companies as you grow? Haritha too, like you're starting a company. Do you think about that in terms of an attribute that you would screen for or manage to like, how important is that going forward?

Trevor Turnbull:

I'll speak to that real quick. Like what Nic's talking about there. I remember 5 years ago, sitting in a role of head of growth for an agency that was part sales part marketing, but mostly meetings. Let's be honest. It was a lot of meetings and we were implementing all kinds of frameworks.

Frameworks like scaling up and, trying to do our huddles and stick to all of the protocols of all of that kind of stuff. And we were experimenting at the time, even too, with ways to collect information during those meetings and then disperse it with action items afterwards and whatnot.

And the technology might have been there, but it wasn't pulled together yet to actually be cohesive in a way that was useful because like you said, we still had our work to do when we walked away from these meetings.

I think it is to your point, John, it is so important that we not only have people that are coming into our businesses that are open and adaptable to learning new technologies, but also learning from others from scratch, right?

What used to work, maybe doesn't work anymore. And what is possible they might not even have a clue about. There's a huge training component to this, and I don't necessarily believe that people even have to have all the skills before they even come in because if a company has their training systems dialed in, they can train people on how to use these tools to improve the efficiencies of their work.

Nic Bryson:

Yeah, the thing that I've looked for over the years in hiring, and I don't think it changes with AI. I think it, it just reinforces the need is looking for that natural curiosity.

If someone's not asking the questions, trying to figure things out in their natural course of going about their work, then, I think it's going to be the person who, we're training how to use a tool down the road from now, once things are more mature, as opposed to the person who's going to go out and explore and try to figure things out just because they're curious.

I feel like that natural curiosity is that X factor.

Brandon Carter:

John, you asked the question, but you also do a lot of hiring here at Promevo. I'm curious to get your take on it.

John Pettit:

Yeah, Nic, I echo your point about curiosity, right? That has been like one of the number one things we look for in terms of people and also try and cultivate within our organization.

And, I think a lot of the AI stuff started off with people being afraid of, are you going to think I'm dumb because I'm using a tool to do my work or I'm speeding things up? And I think we've successfully, I hope at Promevo have gotten to the point where our employees feel like they can do that stuff in the open and they can share with each other how they're being successful with it.

And that we can all win together. It's still keeping the human in the loop. So you're not just phoning it in and copying and pasting, you're using it to accelerate your work, but with some refinement.

But yeah, curiosity is, I think, going to be a differentiator as, as we figure out who's going to help us grow and continue to scale, like it's going to be like, the X factor, like you said,

Brandon Carter:

Haritha, anything you want to add as far as the people that you want to work with and the people that you hire, the people that you're going to bring into Ace4. What are you looking for?

Haritha Yanam:

So for me, because it's like a two man show. We are just starting up, right? So it's 18 hours or 19 hours of work every day.

So I think we need people who are really excited about the product, right? They need to believe the product, and and obviously curiosity is the key, right? You need to experiment so many things in a startup environment.

Nothing is solid, right? You experiment, you fail, you come back, and we experiment again. I think the curiosity is the key.

And then the the the learning, right? You need to consistently learn new things, right? I think that's something I would definitely look into. I don't mind coding. You can learn coding. It's not a big deal, but I think it's that curiosity and then the excitement and then the learning, right?

Keep on learning and upskilling yourself every day. There are so many articles coming up. You open up LinkedIn, you have so many things popped up on your LinkedIn posts, with gen AI. Just keeping up with all these things is hard, but I think, I'm looking, I would be looking for someone who is really interested in learning, keeping upskilling.

Brandon Carter:

I think it's a good transition. We have about 10 minutes left. And I want to spend the last few minutes prognosticating.

We'll start with you, Trevor. Like we, we just, established that people that are going to be working in this field, they need to be curious because it's going to change, and it's going to evolve, and it's going to advance.

In your opinion, like, where do you what are some ideas that you have about what that might look like?

Like, where's this going?

Trevor Turnbull:

Yeah, it's interesting because over the last three years or so, I've dove deeper into personal development myself, as far as, not just looking at my life's purpose and meaning being just my work, but actually my own development as a human, as a dad, as a husband, as a leader.

And I think there's obviously a natural overlap in these things, right? As people become more curious in their work and wanting to learn. They get around other people that are also of that similar mindset and you naturally start to push boundaries of what's possible. And that's what's so exciting about the technology right now is that, I think people are really waking up to the idea that we can't even fathom.

What this world is going to look like, especially from a business environment with all this technology over the next five years, look at how much has changed just in the last couple of years. So it's exciting. And I think intimidating for people as well, but the one constant is a commitment to constant evolution, right? Both personally and professionally.

Brandon Carter:

One of the things that I always talk about with my team is rely on systems where you can, because the brain is it can be deceptive. Like for example our task and project management system, use it because what's in your inbox, you might forget about.

And what, like what someone is like texting you or sending you in a DM or a chat, you might forget about that because that's not a system that you work out of.

So, I think to echo maybe what you're saying here is it's going to become more, more present. And it's going to become not just AI will not just become like a work function or a personal function.

Like a lot of those things are going to come together and like assimilate data in a way that's useful for you as a human being and not just me as an employee or me as, a dad or whatever.

Haritha, what do you think? What's where's this all going? What's it going to look like in the future?

Haritha Yanam:

I think I think there are a lot of interesting tools already in place right now, the Google Copilot, right? So I use that so much in my product development. That's amazing tool and it changed the way we work, right? So I think It's going to be pretty exciting new tools coming in.

And and as Trevor said they need to be balanced between, how we are going to do the personal development and then the professional development and they need to be a balance between that. Otherwise, it might go the other way around.

But I think there is a lot of, a lot of excitement about this. And I'm really looking forward to see and experiment with new tools that are going to come up.

I think it's changing, right? Things, the way we are working is changing. You use ChatGPT quite a lot. Everybody uses ChatGPT these days, so I think there there is a lot of potential for for the next five years.

Brandon Carter:

Sure. Nic, you want to look into your crystal ball? What does it look like for you?

Nic Bryson:

Yeah. It's interesting. I was just reading a survey yesterday an executive summary put out by a Harvard professor who just finished their, uh, survey of executives and leaders around the world at different companies. And the thing that was top on their mind was data management within their digital transformation journeys. And AI was well down the list.

And, I think what's really interesting is that this is such an early time. I think the opportunities that these LLMs and AI tools represent is a differentiating competitive advantage for companies who are early adopters and for companies who will be created to disrupt existing industries.

But I also think about the pace of human change, right? Our ability to build technology moves much faster than humans are actually able to build new habits and adopt and actually change.

And I think while there's a lot of heated excitement about these new technologies, I do think that, over time things will continue to progress at, uh, without anybody necessarily getting, totally left behind per se, it's it's always not as good as it looks, not as bad as it looks. It's that the happy medium,

Brandon Carter:

I think is it safe to say like things are about to get real fast, like they're moving quickly right now, but there's a we're going to see some stuff here in the next 10 years. That's going to be pretty mind blowing. I think, is that, I don't know.

Do you agree?

Nic Bryson:

I think so.

Yeah. I think, the opportunity that AI gives us for to, let's say, disrupt incumbents is is material. And I think that's where, it would be really the most, some of the most interesting stories over the next few years.

Brandon Carter:

John, I'm going to throw it to you, but this will be our final thoughts as we wrap up.

We do have a couple of comments coming in, but I want to go ahead and prompt you in, John, do you want to talk about what as the future?

John Pettit:

Yeah which will definitely be wrong, but I'll start off with, beyond just generative AI, We're seeing some amazing things happening in industries where AI is at the edge.

From crop farming and yields to I took a Waymo on a trip and sat in the back seat and let this car drive me around, surprisingly handled extremely well, or saw a barista making coffee, right, as robot arm, no, no humans there, just AI's making coffees with robotics.

And so like at the edge and in industries that you can't imagine, it's disruptive. But I think also the generative AI integration into the business and knowledge processes is going to be a huge leap forward for people who do any kind of knowledge work.

Again, back to the you're going to think about what AI platform you use, not what productivity platform you use. I think that's where we're going to be. And, it's really interesting about that is we went through like ancient civilization, externalizing knowledge, I think we're still trying to get the last person on the internet somewhere out there in the globe.

So yeah, there's a long tail to it, but AI is going to be the same thing. There's going to be a huge boom and nexus of great productivity growth and companies that take advantage of it. And it'll gradually make its way out to the edge, to the last truck driver with a flip phone, who refuses to have the government spying on them.

We'll get there though. Like we're going to get continued advancement, but yeah, some of the things, Wharton had a conference in September and, people talk about, the enjoyment of their work. Like I personally find that coding again, using Gemini for code or any of these code assist things that it makes me want to write more code. And, interesting that came out of that was When they unplugged the AI and the people who were using AI as like a buddy to do their work for like customer service agents, they were like will the knowledge just drop and the service will go down?

And it's no, it like averaged out to be better than it was before. It's not just like we can't operate anymore without AI, but it does become a companion. I think it does make people faster, better. And it seems to make people have more enjoy enjoyment in the work that they do.

Brandon Carter:

For sure. And, I would even say that like the guy that's hiding out in the woods with a flip phone, like it may not, he may not be like summer summoning an AI bot on his phone, but he's going to benefit it. He's going to benefit from it in other aspects of his life, somewhere out there.

Even in the crops that he's hiding in one of the things that you mentioned there. The field that like that has the shed that he's hiding in like the way that it's cultivated is going to be improved by AI.

Thank you all for that. We have One and a half minutes left. There was a question that came in.

So if we can just do like quick hits. The question is about upscaling and encouraging curiosity in a company, which I think is going to be super important. I've just what are your all what are your best tips to like, hey, I want to give my employees. I want to give them opportunities to be curious. Like what a quick tip that you have for those.

We'll start with Nic.

Nic Bryson:

Yeah. I think one of the things is understanding clear guidelines for that and maybe even some. Suggested tools that you'd want to use. But, we've talked, we've just mentioned the word privacy or security a couple of times in our conversation here, but I think there's a lot of question marks, organizations trying to figure out how they feel about those topics.

And so I think the encouragement for employees needs to come with good reinforcement and trust and, understanding of, those topics, with with those. Maybe suggested tools

Brandon Carter:

Trevor, any thoughts on how do you encourage a curiosity? How do you cultivate that?

Trevor Turnbull:

Yeah, I think it's anchoring it back to what matters to that employee as well, right? It's like you talk about retention is one of the biggest pains for companies of all different sizes, right?

How do you keep good talents around and you do so by painting a picture of what's possible in the future with that company so that they can continue to grow so I think if there's a way to not only, look at this from a business application of here are these tools and this is how we would like you to use them to upskill and motivate yourself to improve.

You also show that person how this will actually advance them in their career and their lives. And, even just in their mindfulness.

There's so many advances in AI and creating custom affirmations and meditations and things that just help people ground in a seemingly chaotic, busy world.

So there's a lot of overlap in these things. It's really exciting. If employers look at it that way,

Brandon Carter:

Haritha, final word.

Haritha Yanam:

Yeah. I think one thing adding this AI literacy helping employees upskill their AI literacy, it gets a pretty intimidating for folks who are not from tech, AI.

So I think adding this regular training sessions, getting that excitement up going what is prompting, what is prompt engineering, start from scratch. And then helping them, giving them some role based access to experiment with the tools and then making sure you give the guidelines that you can, what kind of data that can they use with these tools, right? You cannot use sensitive data. What is proprietary data?

I think getting that AI literacy in and then giving them quick guidance on how to experiment with the tools. Builds that curiosity. So I think that would be good.

Brandon Carter:

Awesome thoughts. Everyone, thank you for joining. Much appreciated. That was a lot.

We went through a ton of stuff in an hour. We actually went over by a couple of minutes. But Haritha, Trevor, Nic, John. Much appreciated.

To all of you out there viewing. You were quiet, but I saw you up there. I know you're watching. We thank you for joining us.

And yeah, if you ever want to get into a development or try Gemini, reach out to Promevo.

We're happy to help with that. It's a big part of what we do. And as always, go to promevo.com and sign up for our newsletter. And you'll hear about more webinars like these.

We have a ton more content that we're developing just to help people navigate this AI journey that we're all going on.

So with that, thank you, everybody. Have a great Tuesday and we'll see you all soon.