The Future of Government Services: Enhancing Public Engagement through Natural Language Search Tools

In a live stream by Avèro Advisors, the founder and CEO discussed the importance of artificial intelligence (AI) and its integration into the public sector. He highlighted AI's potential in improving paper-heavy and process-intensive processes. He also mentioned the use of AI in generating documentation, reading documentation, and ensuring compliance with regulations. The CEO emphasized that while AI may change job roles, it will not eliminate jobs as human verification is still needed. He also discussed the potential of AI in aiding crisis communication or emergency responses. The CEO encouraged public sector entities to envision how AI can improve operations and explore AI-enabled ERP solutions and natural language reporting.

Stay up to date on industry trends!

Download our free eBook:
➡ https://bit.ly/2023techguide

Learn More About Avero:
➡ https://www.averoadvisors.com

Connect With Us:
LinkedIn: https://www.linkedin.com/company/averoadvisors
Facebook: https://www.facebook.com/averoadvisors
Instagram: https://www.instagram.com/averoadvisors
TikTok: https://www.tiktok.com/@averoadvisors

Connect With AV:
LinkedIn: https://www.linkedin.com/in/verekar

(865) 415-3848 | info@averoadvisors.com

Transcript

Megan (00:25):

Hello everyone. Happy Monday and welcome to the Avero Advisors live stream. We go live once a week, and today I have Mr. , who is our founder and CEO with me. And we are going to discuss the importance of AI or artificial intelligence as you all may know it, and how those tools can be integrated into the public sector. So, hey ab, how are you today?
AV (00:51):

Hey, doing good. Excuse my voice. I'm just getting over little bit of a crud medical term,
(01:01)
But thank you for this important topic. I think I did a talk a couple of weeks ago with the East Tennessee Purchasing Association with our favorite people purchasing agents and talked for more than an hour on the applications of AI and how it can improve their lives today. And generally, there is a sense of excitement, fear, and curiosity about the whole thing. So since then, I've had a couple of meetings with various departments that are paper heavy, that are paper intensive, process heavy, that could use AI in many formats. And what I also said to them is no one knows how this gets supplied. Many people are experimenting with everything that's out there. A lot can be done on GPT today. Every day OpenAI releases new functionalities. So I encourage anyone to just go spend the 20 bucks, get GPD four and use the plugins. And what they most recently came out with is custom gpt. You can create your own gpt. So I created maybe five over the weekend just trying to get more expertise into the chatbot, right? It's not as generic anymore. So anyway, went off on a tangent on you there. Megan (02:26):

Sure. No, no, no, you're fine. And before we delve into things, guys, if you're joining us today, please comment where you're joining from and feel free to ask questions or post your thoughts there. And I didn't introduce myself. I'm Megan Seton, I'm the business development manager with the Vero. And to get started, let's talk about the current state of public sector communication. So how do current communication methods, public services, meet or fall short of public needs? AV (03:00):

Most cities, counties, any government agency, they expect you to come into the office or call the line to get information. And we all know how clunky websites tend to be for information dissemination. COVID was kind of a boon for that sort of function to happen because you expected to work from home. You were expected to do business remotely. So a lot of our clients were able to put in the practices and the tools for citizens and consumers to communicate remotely with the city hall or county courthouses. That kind of quickly went away once covid restrictions were lifted. Right? A good example is how our RFP submitted. We submitted hundreds of RFPs over the last two years, and in the covid year was all electronic. It was on a portal. Lots of people went out and got portals like ion wave or vendor registry, and then they were accepting bids online.
(04:05)
And for some reason after Covid that went away too. We're now sending bids, several copies of printed bids and proposals to our clients or prospects. So communication outward facing is still challenging. I think there's a lot of hope there because of AI enabled chat bots because how often do you click on the chat bot button to get the same information? The bot will tell you that, Hey, call this number. Or it'll say, go on our website. I am on your website. But all of that's going to change with AI enabled bots that you are starting to see this where you don't see the difference between a chat bot and a human if it's done well. So that's one area of communication. The second is internal. Right. When you talk about communication within the organization, what are you mostly looking for? You're looking for financial reports, you're looking for trends, you're looking for audit information.
(05:04)
How is that being disseminated within your organization? One, we see it every day. We see it today, we're doing business on Excel sheets primarily. We have a new client that is looking to us to overhaul their budgeting processes, and they said because of how big this file was, it finally broke and they lost some data. Not all of it, but all of that's about to change because of AI and natural language processing. Again, how those Lego bricks are put together is up to consultants like us and software developers, but the need is absolutely urgent and out there that says we need better reporting systems. We can't wait on it. Or someone that knows how to do Crystal Reports or Power BI to spend six months to get us a very simple report and no fault of theirs, right? It's a very customized, specific thing that the CFO is looking for that may not be out of the box.
(06:06)
So someone has to spend the time to understand what that means, where that data set exists, how to build it together, how to present it, and then have the feedback loop. So what I have played around with so far in GPD four and all of the plugins is that should become a lot more easy. And Bill Gates not, will Gates, bill Gates, the real Gates had an article on LinkedIn yesterday that talked about how AI is about to change the lives of people in the next five years and he said it 30 years ago, and now we have the tools that are going to help us get to that, get in that direction. Again, natural language reporting. I know you want to ask me that question, but I think it's the most exciting thing to come out of this AI revolution for especially our CFO clients. They're struggling. Megan (07:01):

Yes, absolutely. Let's go back to the presentation that you and I gave a few weeks ago to the procurement team. I do want to get to natural language reporting, but what are some of the more simpler ways that AI can be used, especially for procurement in the public sector? AV (07:20):

I think drafting documents, because purchasing agents are, their primary job is to make sure that we're buying the right things for the right amount and that the contracting is done right, that the risk is managed. So all of this involves a lot of writing, and if you are not an expert in a specific thing that, let's say IT department or public works wants you to buy for them, then you're in trouble because you don't know how to put that scope of work together. Now, this impedes a little bit on what we do, but in a good way because not only can purchasing departments come up with scopes of work and requirements at a very basic level, they can also use GPT or any other AI like copilot was just released by Microsoft for general use, which uses the same engine on the background, right? Open AI's AI machine, and you can ask your questions without the fear of someone judging you for not understanding the basics, right?
(08:22)
Go ahead and ask those questions. Use GPT. Highly recommend getting the four version. Now they have GPT Turbo, which is supposedly even better. That's again, not out for general consumption, but it is in the background being used by GPT four. So using those tools to generate documentation, to have it read documentation, compare notes, make sure that your RFP that you're just putting out or the proposal you receive or the contract negotiating meets things like Davis Bacon Act. It meets the 1957 purchasing Act in Tennessee. So you don't have to be well-versed in those things now to make sure that this document you're looking at is actually compliant with those very important regulations. So those are some very basic ways you can really leverage GPT and other AI tools that let you upload documents. And to my knowledge right now, it's just G PT four that does that. Megan (09:27):

So a lot of the comments around AI that we received during that presentation revolved around, Hey, will my job go away? Because essentially as a procurement officer, that's what I'm doing. I'm writing. And you mentioned that earlier. So what was your advice and what is your advice to our procurement people out there who are fearing or anybody who's fearing for the AI is going to take over their job role? AV (09:56):

Your job as you know it today is going away. Yeah. Does that mean you're going to be out of work? Probably not, because you still need, at least in this iteration of the AI revolution, you still need humans to verify that the AI is not hallucinating or lying to you or making up numbers. So that's one basic way that you can still be in charge. The other thing is you still need to tell this machine what you want it to do and give it context and give it meaning, and then it learns, but it's nowhere close to being at human intelligence as far as creativity is concerned. That might change too, but again, we need to look at this as an incredible tool. What a great time to be alive. I wish we had at 20 years ago when I was starting out in consulting.
(10:49)
It's wild. So take every opportunity to use it and don't be afraid because yeah, your job is changing. It will change, it will evolve, but it's definitely not going away because we still need people in those very important positions like purchasing agents in Tennessee are very important, powerful people. I don't see GPT replacing anyone, but it's definitely going to make your life easy. And one of the ways that things we talked about a couple of weeks ago in Gatlinburg was that you are facing a labor shortage everywhere, right? It may be about to change, who knows, but when purchasing departments, when we go into any office, they're like, well, we are short three people and we are struggling and we have this contract deadline coming up and we don't know what to do. Simple things like coming up with renewal notices, right? If you don't have a good contract management module within your ERP system, you're left doing this manually. And to do that for hundreds of contracts is very inefficient. So we're working with several clients to figure out how to automate all of this while keeping the quality assurance in place and making sure that operations keep going on an efficient basis. Megan (12:11):

Yes. We do have a question in the chat. Thank you, Mackensie. She says, what would be the optimal scenario for integrating AI into one of the basic fundamental government services, and what would that change?

AV (12:26):

I think purchasing is a great example to stick to that topic a little more because again, it's very paper driven. It's very, even if it's not paper driven, it's very document driven, right? You need to interface with the department that's asking you for a certain thing. You need to know exactly what they're asking for so you can research your options. If it comes to putting an RFP out, it's also very document driven. You need to have the right requirements, the right language, the right drawings and what have you. That goes into an RFP process. And then when you get bids in, you need to be able to understand what they're saying. Yes, you'll have your selecting departments to help you with that. But again, as a purchasing agent or director, it's on you to understand what exactly is being purchased and if this price is valid or applicable.
(13:21)
The other thing is, when it comes to contracting, this is another scenario we were in last week, how do you deal with the law department in a good way? This client we were in, it's a county, they have an external legal department, meaning they outsource to a law firm. So when you send a contract that you think is perfect and you need to buy this thing today, and they sit on it for a while and then they'll send you a paragraph that needs to be changed because it doesn't comply with something, you could leverage AI and GPT to send them a clean version, meaning tell GPT to put in those clauses that apply to a certain industry, to a certain state, to a certain county that might make it cleaner, better, faster for your legal department to turn things around quickly. Not saying it's going to do their job, but

Megan (14:18):

It helps streamline their day-to-day, right? It gives them more time during the day to spend time thinking creatively and not so much time typing out changes. So I agree with that. It's helped streamline my day-to-day job. So in regard to natural language reporting, so what does that even mean? What is your understanding of it and how does it fit into the broader AI landscape?

AV (14:47):

Well, it really means asking questions like a human being, natural language. Like you ask Siri, Hey, what's the weather today? It's going to tell you how does it happen in the background? There's code being transmitted back and forth that's translating human language into computer language and then back to you in human language. It's not an easy thing to do, but Apple's been at it since, I don't know, 2009, 10, and Siri is not perfect, not even close, but if you have the GPT app on your phone, it has voice instructions now, so you can talk to it in a natural way. So imagine that interacting with your data, right? You plug that into, and I'm saying that because it's literally open AI's engine or Google's or whoever else's. You plug that into your backend of your ERP and it interacts with your data. It's very similar to plugins and PDF files that you can now chat with on GPT.
(15:57)
So your GPT version is going to go get you the data that you want without you having to program anything. So you say things like, show me a graph of how many sweatshirts we've bought for public works last year, or Tell me which vendor we've spent the most money with, or how much money do I have on a certain po? And traditionally, you've had to go to your IT department to get them to write reports for you for something as basic as that. And then you have more complicated reports that organizations have multiple report writers that do nothing but try to decode what it is the regulators want, what it is the executives want, and then try to create that report out of a database with natural language reporting. And I'm oversimplifying this, but really that's the vision that you can now have dashboards and reports that you chat with. Like you chat with GPT, you would have GPD plugged into your database, and you would ask it questions as simple or as difficult as you want, and it should be able to deliver those to you.

Megan (17:12):

Go ahead.

AV (17:13):

I did that yesterday and I'll show you later. We are working on our annual retreat and financial reports and stuff, and GPT has been a huge help. It's sort of data for me. It's created graphs, it's done all of these things. That would've taken me days. Megan (17:36):

Yes. And it took me all day on Friday to come up with the very small baby report that I came up with in Excel. So I can't wait for you to teach me the AI way. Can you share examples of how natural language reporting has been used or is being used in the public sector right now? Do you have any examples?

AV (17:57):

There's some, I mean, it's not widely used yet, and natural language reporting has, it's not a new thing. It's been around if you've used Power bi, power BI has had the ability, so if you have a spreadsheet with a big dataset, go into Excel and find the button that says Analyze data. Once you click that button, it'll come up with a box that says, ask me any question about this data. That's natural language reporting. Of course, it's not perfect because it's a little clunky compared to what AI can do, but it's been around for a while. We've seen chatbots that have been applied to cities websites that quickly give answers. You can go on a website today and search, but you know what happens then when you search for a term, you've got 20 pages of links and you don't know which one is the right one.
(18:57)
If you apply an AI engine to this, you can now find the exact specific thing you're looking for. We're trying to help a city improve their permitting processes, and one of the things we're implementing is an AI search bot that if you're trying to apply for a permit to put a deck on your house, you can just go in and say, I want to apply for a deck permit. How do I go about this? It's going to send you to a page that gives you the exact results step by step with links so you're not lost trying to find a specific permit type. So lots of cool ways it's being used right now. It's not very publicly available, but I'm sure you can go down a big rabbit hole if you start looking for things that are practically being used.

Megan (19:50):

What are the primary obstacles in adopting natural language reporting or AI tools?

AV (19:57):

I think an understanding of the tools available is one, it is vast, right? You've got open ai, you've got Google Board, you've got copilot from Microsoft, all of these things. And then if you're a little bit savvy with Microsoft's Azure environment, you've got the open AI APIs available to you, but you need to be a little tech savvy to understand how these pieces fit together. But I'll tell you what you should be doing as ACFO, as ACIO, that is maybe not as tech savvy. You should be thinking about ways that you think AI can improve your lives, have a vision, because people like us or developers, we don't understand your world so much. There's going to be a time where the tool sets and you have the problems, and there's a huge gap in between because they don't seem to meet. So we need to have a lot of dialogue about what's possible, what they would like to see, and then bridge that gap by actually doing some proofs of concept and showing the people that need this technology to work for them, how it can be applied practically.
(21:10)
I think imagination is the best thing to do right now because the tools exist and tell us what you want this to do. We have hopefully a new client that has been talking to us about robotic process automation. Now, RPA has been around for a long time. It's not a new thing. It's just becoming mainstream as AI is picking up. So combine AI with RPA, and it's going to be dynamite for those that understand how it can be used and that you can't just throw warm bodies at business processes anymore, that you have a lot of great options available to you. So sky's the limit.

Megan (21:58):

Yes, and I think it could play a huge, huge role in aiding crisis communication or emergency responses. What are your thoughts on that? How could AI be utilized in that way?

AV (22:11):

I think maybe some predictive communication, right? I'm sure that's what they're doing in Islam right now. They've evacuated the whole town because they think a volcano is about to erupt. And again, again, that data set has been available for a while. What's changed is the speed at which you can analyze and predict things. And then the second thing is disaster relief. Prior to this, even now, if there's a tornado, you need to go out there, take pictures, see what houses are destroyed, what the definition of FEMA for destruction is, and at what level they're going to get payment as relief. So all of that takes a while because you have to send humans out there to take pictures and record things in the FEMA forms and send it back to the state, and the state communicates with fema. All of that can happen seamlessly. Now with ai, you can just go take pictures and have that picture read into a database and say, okay, this classifies as a completely destroyed house. This one does as partial destruction, send that to the database and quickly have those claims filed so people don't have to wait for months to get their relief checks.

Megan (23:33):

How do you, looking ahead, how do you foresee the evolution of AI in public sector communication and what new opportunities or challenges may arise in the, AV (23:46):

I think public sector is in a uniquely advantageous place because one, most things they do is in the public domain. So you're not dealing with feeding the AI machine classified information or information that should not be public. Of course, there is information like that in the public sector, but most of it is not. Your RFPs, your contracts, financials, all public information. The second thing is, and I've always maintained this because the public sector tends to be slightly behind the curve when it comes to technology, it's to their advantage because now you can leapfrog bad technology. So people or agencies that are still using ASS four hundreds or they're using green screens and are looking for a more modern solution can now in the next two years or next year, look forward to AI enabled ERP solutions, AI enabled reporting, natural language reporting, and it's going to make life so much easier on the implementation side and also on using these new systems.
(25:00)
So lots of great ways to use. Look, our practice relies on creating a lot of documentation reports and requirements, and we've embraced AI enabled consulting completely. It's taken the timeline from nine months to get an RFP on the street to three to four months, because we are now much faster at doing things, and it's not just GPT. We have several other tools that we use to, we've shortened the timeline, and therefore the money it takes to get you to an RFP. Traditionally, we would have an army of consultants sitting in your offices just doing business processes and requirements. We still do that. We still pay that attention, but cranking our deliverables and outputs has become easier. Yes.

Megan (25:49):

Specifically around it, strategic planning, right? The actual plan, we can turn that around pretty quickly now once we gather the information,

AV (26:00):

And as long as the data's available, as long as we are able to feed it to the ai. Of course, like with IT strategic plan and cybersecurity assessments, we don't want to do that. But yes, the more mundane things have become a lot quicker. So I think the timelines being reduced is the biggest win for our clients.

Megan (26:24):

Absolutely. av, I know you've done a ton of training around all things ai, so where can our viewers go and learn more about ai? Where can they read more about it? Who can they contact? What courses can they go out and take?

AV (26:41):

So I think the first thing to do is just go experiment. Really, no one can teach you better than you can teach yourself, but there's a lot of resources on YouTube. It's up to you to go down that rabbit hole. But we are available. We're on every social platform on our website aware advisors.com. You can get in touch and see how this revolution can help you and your agency.

Megan (27:10):

Yes, absolutely. Well, thank you guys for joining us today, av. I don't have any more questions, but if you have any final thoughts, I'll let you give those now. AV (27:19):

No, I think we're good. Thank you for joining us today and look forward to next time. Thanks, Megan.

Megan (27:24):

Yeah, absolutely. Again, as AV mentioned, he kind of stole my thunder, but we are on every social media platform. We're very easily accessible. Please reach out if you have questions or concerns, or if there's something that you would like us to speak about next week. We are going live on the 22nd around thankfulness and what we're thankful for and what we're grateful for here at Aero. So thank you for joining us today and look forward to seeing you guys next week. Thank you. Thanks. Bye.