How NetSuite uses AI and how the NetSuite Connector is a Game Changer for accountants with Joe and Ranga
In this episode of Future Finance, hosts Paul Barnhurst and Glenn Hopper sit down with NetSuite’s Joe Friedman and Ranga Bodla to explore how AI is transforming enterprise resource planning (ERP) systems. With decades of combined experience in enterprise software, Joe and Ranga discuss how NetSuite is innovating ERP through AI-native capabilities, open platforms, and real-time automation. The conversation spans AI agent adoption, data clarity, and how finance teams can start harnessing AI today without major overhauls.
Joe Friedman is the Senior Director of AI Innovation at Oracle NetSuite, driving efforts to integrate practical AI across finance and operations. With over 15 years at NetSuite, Joe focuses on converting AI into measurable productivity gains. Ranga Bodla is the Vice President of Field Engagement and Marketing at Oracle NetSuite, with more than 20 years in enterprise software. He leads industry strategy across verticals, helping businesses convert innovation into performance.
In this episode, you will discover:
Why NetSuite believes ERP is the core of AI transformation in finance
How the new MCP connector integrates AI agents like Claude and ChatGPT
Practical AI use cases: from variance analysis to journal entry creation
Why clean ERP data is essential for effective AI outcomes
How companies can start AI initiatives without high risk or cost
Joe Friedman and Ranga Bodla offer a compelling look at how NetSuite is leading the charge in making AI practical and accessible within ERP systems. Their insights show that businesses don’t need massive overhauls to start seeing value from AI, just clean data and the right tools. As ERP evolves, NetSuite’s hybrid approach positions it to stay ahead. This episode is a must-listen for anyone navigating the intersection of finance, tech, and AI.
Join hosts Glenn and Paul as they unravel the complexities of AI in finance:
Follow Joe:
LinkedIn: https://www.linkedin.com/in/friedman-ns/
Follow Ranga:
LinkedIn: https://www.linkedin.com/in/ranga-bodla-bb45b/
Follow Glenn:
LinkedIn: https://www.linkedin.com/in/gbhopperiii
Follow Paul:
LinkedIn - https://www.linkedin.com/in/thefpandaguy
Follow QFlow.AI:
Website - https://bit.ly/4i1Ekjg
Future Finance is sponsored by QFlow.ai, the strategic finance platform solving the toughest part of planning and analysis: B2B revenue. Align sales, marketing, and finance, speed up decision-making, and lock in accountability with QFlow.ai.
Stay tuned for a deeper understanding of how AI is shaping the future of finance and what it means for businesses and individuals alike.
In Today’s Episode:
[04:15] - Competing with New ERP Platforms
[09:42] - What is MCP & AI Everyday?
[13:47] - AI Use in Daily Finance Tasks
[20:18] - Why Many AI Projects Fail
[27:14] - How to Start with AI in NetSuite
[30:25] - Building AI Workflows Without Engineers
[35:19] - Future of ERP: Embedded AI vs. External Agents
[39:33] - The Most Human Thing About AI
[43:49] - Closing Thoughts & Takeaways
Full Show Transcript:
[00:01:41] Co-host: Glenn Hopper: Welcome to Future Finance. I'm Glenn Hopper, along with my esteemed co-host, the Reverend Doctor Paul Barnhurst. How are you doing, Paul?
[00:01:49] Co-host: Glenn Hopper : Reverend, today I'm doing good. Now.
[00:01:53] Co-host: Glenn Hopper: Our guests today are Ranga Bodla and Joe Friedman, and we'll introduce them in a minute. But welcome, guys.
[00:01:59] Guest 1: Ranga Bodla: Thanks for having us.
[00:02:00] Guest 2: Joe Friedman: Yeah, thanks. Excited to be here.
[00:02:02] Co-host: Glenn Hopper: We've had all these new kids on the block in the ERP space on lately, so we thought we would. It would be a great time to go to the mothership and talk to you guys at NetSuite. So today on Future Finance, we're going to look at how NetSuite is reimagining the ERP in the age of AI. You know, from conversational interfaces and embedded agents to continuous monitoring and intelligent automation, we want to talk about NetSuite next and the promise to make enterprise systems more connected, contextual and collaborative. So a little bit more intro on our guests. Two people at the center of that transformation are Joe Friedman. Joe is the senior director of AI Innovation at Oracle NetSuite, leading efforts to embed practical, measurable AI across finance and operations. Over his 13 years with NetSuite, he's held roles spanning the Finance Center of Excellence, Chief of Staff and Solutions Consulting, building a reputation for tuning intelligent automation into real productivity gains. And Ranga is Vice president of field engagement and marketing for Oracle NetSuite global business unit with more than 20 years in enterprise software at IBM, Hyperion, SAP and NetSuite, he drives industry strategy across manufacturing, software, services and distribution, helping organizations turn technology innovation into performance results. And Paul, I think that's why we don't let me do the intros more often. I can't do anything like quickly. I'm halfway through the podcast and we've just got through the intro.
[00:03:24] Host: Paul Barnhurst: That's all right, Glenn. We won't hold it against you. At least I won't. They may.
[00:03:28] Co-host: Glenn Hopper : Maybe I should stop talking.
[00:03:30] Guest 2: Joe Friedman: I won't, I won't hold it against you. I will say, though, that I now have 15 years at NetSuite, so you know, so it's not a competition between Joe and I, but you had a couple of NetSuite veterans here in terms of our focus here.
[00:03:42] Co-host: Glenn Hopper : Love it, love it. And Paul, I guess I should turn the questions over to you, right. And stop talking myself. So people aren't just hearing my voice. There's four of us here.
[00:03:48] Guest 1: Ranga Bodla: I need to be quiet. It's a nice voice. You've got a great voice for podcasting. That. That puppy's smooth.
[00:03:54] Host: Paul Barnhurst: Glenn. Glenn does have a good voice. That's why I. I let him out of the room occasionally.
[00:03:58] Guest 1: Ranga Bodla: Because Glenn rather sorry, Glenn.
[00:04:02] Host: Paul Barnhurst: But, you know, hopefully I have a decent voice as well. I'm not hopefully chopped liver, but I'm not quite Glenn.
[00:04:08] Co-host: Glenn Hopper : And a way better beard, by the way.
[00:04:09] Host: Paul Barnhurst: Yeah, at least I have a better beard, so I do have that going for me. All right, we'll go ahead and start here. You know, first question is kind of ask is right. We all know NetSuite, especially in the mid-market, has been the dominant player for years. Right. For many companies often felt like the obvious choice. But now we're seeing a wave of new ERP platforms raising serious funding to take on the incumbents. Last number I saw top six tools I can think of have raised 640 or no. Yeah, 680 million or something. Right. The number just keeps going up. Where does NetSuite sit from your guys's perspective in this changing landscape? And what do you think is driving the renewed interest in all these new ERP tools? I'd love to kind of get your perspective as someone who's been at the heart of this for so long And, Joe, either one of you can take that.
[00:05:00] Guest 2: Joe Friedman: I'll start. And I'm sure Joe will have a lot of perspective here as well. I mean, look, the reality is there's been, you know, the death of ERP as a conversation for years, decades. And, you know, I think the reality is the ERP system is often the central nervous system of an organization, really powers where an organization can go, can operate. And so you had a bunch of new folks who are out there looking and saying, you know what? We think there's an opportunity to drive some innovation into the market. We still think ERP is a category that's important. And so we're going to invest here. Now I don't think it's a bad thing. You know a little bit of competition is always good. I think the reality which I'll make though is we see over and over again the competition isn't the competition. The competition is just the status quo. People don't want to, you know that. They've got a system, they've got pain, but they're hesitant to make a move. And so, you know, for us we're going to continue to innovate. It's been something we've been focused on as an organization is keep on moving the ball forward. You saw us at a couple of weeks ago at Sweet World. We delivered a ton of new announcements, really exciting announcements, which I think we'll get into here, as Glenn brought up in his intro. But we see the we're just getting started in terms of the innovation that we're doing. It's something we've always been passionate about, is continuing to deliver new functionality to our customers. That really helped them solve their business problems, and that's where our focus is. We're really not trying to pay attention to the amount of money that's being raised or, you know, that side of things we're focused on is what do our customers need, and how can we leverage both our market position and frankly, the power of Oracle to continue to dominate in the space?
[00:06:42] Host: Paul Barnhurst: Sure. Yeah. Oracle's a big name for in the space for sure. And we know when you talk dollars and capitalization there's not many bigger.
[00:06:50] Guest 2: Joe Friedman: Absolutely. And you know you know I mentioned I've been here 15 years you know six years prior to acquisition. And then of course the other 9 or 10, I guess, if I do the math right. Yeah. Nine. You know, after the acquisition and Oracle, you know, we're they're investing in NetSuite. We're able to take advantage of all of that that Oracle investment, all the stuff they're doing in AI. We get you know, all the stuff that you hear about OCI and Oracle AI. Those are all things that NetSuite gets to take advantage of. And you know, it's a huge, huge thing that we're not having to spend tens of millions of dollars, hundreds of millions of dollars, billions of dollars. Oracle's doing all that that we get to take advantage of. Joe, do you want to add anything to that? Just from what you see?
[00:07:29] Guest 1: Ranga Bodla: Yeah. I mean, look, I think any AI meeting or presentation you'll ever see, they're going to say the best AI is built on the best data. And I think ERP being sort of the source of truth, I think it's no surprising that we're seeing all of this renewed interest in ERP. And I think a lot of people look at, you know, implementing an ERP system as that opportunity to clean their data, get good data in. You know, these startups are bringing a lot of energy into the space, and it's no doubt that they've done a good job deploying a lot of their newly raised capital on marketing. But the fact is, NetSuite is battle tested and trusted, right? We've got almost 45,000 customers, over 600 public companies. We are proving every day, especially, you know, with our new AI connector and using NetSuite with cloud, that the data model that NetSuite has is incredible at really driving accurate AI outcomes as Ranga mentioned. We get to leverage all of Oracle's top down investment in cutting edge technology, and we've really taken that investment, and we've retooled our entire technology stack from the foundation up, whether it's, you know, going on OCI data centers, you know, on the latest and greatest database technology, all the UI technology that we're leveraging in NetSuite.
[00:08:43] Guest 1: Ranga Bodla: Next. And honestly, a lot of the things that were always our strengths, right. The fact that we've got an open platform, that we've got a whole suite of records from CRM to finance the purchasing in the sweet. It's sort of AI, sort of accentuated our strengths and almost mitigated a little bit of our weaknesses. Right? You know, some of our when you use a cloud or ChatGPT and you connect it to NetSuite and you have access to all this data, you can almost make magic happen. And now, you know, maybe some of the feature depth or gaps between some of the industry leading CRMs and maybe our CRM, it's mitigated a little bit. And just the fact that we've got those CRM records and we've got the configurability of the records, all of those strains are really accentuated now. So I think, you know, I'm a capitalist. I think it's great that there's a lot of new competition in the space, but I think NetSuite is incredibly well positioned, and we're really leveraging a lot of this Oracle technology to sort of leap ahead in the next. I think you're going to see that really coming next couple of months.
[00:09:42] Co-host: Glenn Hopper : Joe, and I'd love to drill down on that. I love that you brought up the MCP connector, because Ranga and I talked before the show about the AI everyday outside in and inside out. I'd love to if one of you guys could kind of unpack that for me and break down what you mean by that.
[00:09:56] Guest 1: Ranga Bodla: Yeah. So, you know, one, we've always been sort of committed to having a really open platform, right? You've always been able to extend the capability of the suite. Right? We've had suite script, we've had suites flow, we've got this huge ecosystem of partners, and AI is sort of just blown the roof off of now what you're capable of by having this really open platform. And so we made a commitment to making NetSuite data, you know, and tools extremely accessible from these industry leading large language models like cloud, like ChatGPT. We're going to be following that with, you know, all the rest of them, Gemini, etc.. And, you know, I've been living in cloud and NetSuite the last two months and it's incredible what you can accomplish. Right. One another sort of advantage we have is NetSuite has been on the internet since 1997. Every user guide, every feature, every piece of information that's ever been put out about NetSuite in the last 20 or 30 years, almost is on the internet, the same internet that these models were trained on. So these models are just got an incredible wealth of knowledge about NetSuite. And now you can load them with context and instructions, and you connect them directly into our data model. And you can do things like, I mean, just get deep insights into reports and you can drive, create, write transactions. I mean really access to all the reports, access to all the searches, access to all the records, read, write access, totally role based. Right? So it's sort of just like this idea that we're going to allow our customers, right? You know, you pick your LLM, you connect it to NetSuite, you're going to be able to make magic happen.
[00:11:30] Guest 1: Ranga Bodla: But then we're also right. We're also bringing that AI first or AI native experience directly to our UI. Right. It's a new look and feel of NetSuite, but it's going to be a whole new way of interacting with NetSuite directly in the application via the Oracle User Interface will allow you to converse in natural language, create records in natural language. Get those, you know, narrative insights directly in the sweet. So really, you know, we call it a bunch of things, but it really is your way. You want to log into NetSuite and you want to have that true first experience. We're going to be delivering that via NetSuite next. But today, right now you can connect NetSuite to these industry leading language models. And you can really do some incredible things. And it's just the beginning, right? I think our open platform is just going to enable, you know, ChatGPT just announced it. You know, the whole agent workflow builder, you know, that's going to work great with NetSuite, you know, and and the sky, I think is the limit on what you're going to be able to do with the sophistication of our data model, the openness of our platform and the ability to connect it to these, you know, extremely, extremely smart NetSuite genius consultants named Claude and ChatGPT, etc..
[00:12:41] Guest 2: Joe Friedman: I'll just pile on to what Joe said, you know, Glenn and we talked a little bit, or you've written a lot about this whole concept of shadow AI and the fact that, you know, a lot of these big enterprise AI projects are failing. It's like they're failing on one end. But then shadow AI is like doing really well. And I think it, you know, this, it's one of the reasons we're really enabling what we're doing with the connector. And why I think MCP is so powerful is the reality is we don't want people to just go out and spend all this money on creating, like some bespoke AI project. We'd rather them use these off the shelf tools if that's what works for them. And then at the same time, with what we're doing with NetSuite, it's about making it part of their process, part of their day to day, as opposed to, oh man, I gotta learn something else new. I'm already doing this. I'm already within context of the application. Here's how I want to use it, and really thinking about AI that way. So it is really, as Joe said, you know, the AI, not just AI your way AI every day and making it so you're not having to like, oh man, I got to figure out all this other stuff, this new stuff. No, we want to ease that burden as opposed to make it a new burden for you.
[00:13:47] Co-host: Glenn Hopper : I think that's really interesting, especially in light of It's been over a year now, but remember when Satya Nadella basically rang the death knell for SaaS tools across the board? He said, you know, AI is just going to wipe them all out because they're all just UIs sitting on top of tables. But truthfully, there's the comfort that people have being in a system. But I love that the approach to okay, within the software, you can interact with AI, but also you can bring your own models and interact with it however you want. But it is still, I guess, to nadella's point. It still is, you know, a different way through MCP to connect to those tables and get information out of it. But I like the approach of two doors in.
[00:14:24] Guest 1: Ranga Bodla: It's not just get information out of it, it's analyze that information and then push information back into it. When you create a cruel journal. Entries from Claude. Right. Created by Claude AI checkbox checked, full audit trail. You know, talk about human, you know user. You know, humans in the loop. You know, Claude is really just acting as an extension of the user that it's, you know, it's connected to the permissions and the restrictions that are given to that user. So when it creates a transaction, it can see that Joe Friedman created that transaction via the APIs, via Claude. Full audit trail on that transaction. I can then have a whole search on transactions created by AI. Who did them. I can audit those. So I think there's a real benefit to having that source of truth, right. Having that full auditability, verifiability, observability, but then being able to use this like expert genius assistant to help you do your job, analyze these huge data sets and then push data back in. And I think now it's just going to be about optimizing your these, you know, call it what you will call it your data tables or call it your NetSuite account, but optimizing it for use with these tools to get the most out of it, to get your people the most productive. And I'm just, you know, it's really the lightbulb. Every day I just am like, oh man, now what? We're going to be able to do this and this and this. And it's exciting. It really is.
[00:15:42] Host: Paul Barnhurst: Kind of speaking to that. That leads me to the next question here. We want to ask and just kind of Right. The whole ask Oracle AI agents how is that really changing right now? Where are ways you're seeing the daily experience change for finance teams? Like are you seeing them use it for reconciliations or are they using it for journal entries today. You know variance analysis. Love to know some kind of practical use cases that you're seeing at this point and how you're seeing them be used.
[00:16:10] Guest 2: Joe Friedman: I mean, you know, one of the big practical things, you know, Joe did talk a lot about insights. You know, we asked people what's their key pain point? Not necessarily not users in NetSuite, just in general. What's key pain point for finance users? Financial reporting is still like that's right up there. That's like the number one almost by far, and by the largest percentage of what people struggle with. They really struggle to get financial reporting and visibility into what's going on in their business. And so that's one of those insights that we're seeing get enabled with the MCP connector as an example. You know, some of those harder problems they were trying to solve. The data's there. But then like really understanding and getting that perspective. These brilliant tools like Claude and ChatGPT are helping them solve that problem. But we're also seeing it on, you know, with the other analytical tools that we provide within NetSuite that is really still one of the big areas that people really want help. They need help. And they're looking to to understand that. And Joe can probably give you some more perspective as well from his side as well.
[00:17:09] Guest 1: Ranga Bodla: Yeah, I mean, everything from financial analysis, deep, you know, flux analysis, creating board reports. And then, you know, you get it's sort of like, you know, one leads to the other, you analyze your financials and then you see, hey, my professional fees this month were, you know, went up x percent or my cost of goods sold, benchmarked against, you know, my other companies and my is too high. Well, let's take a deep dive into that and then you can go, you know, and do a full analysis of your cost of goods sold. And you can. But then beyond that, analyzing for anomalies. Right. You miss a recurring transaction. You can create the accruals for the recurring current transaction exception management, the huge piece of this. But then all these sort of repetitive tasks managing your receivables, managing your payables and then even operationally, right, you know, you can now drag and drop a couple receipts in and create expense reports. You can, you know, read documents and create the downstream transactions. And it's just getting more and more sophisticated. Right. Obviously with NetSuite next and ask Oracle and some of the things we're going to be doing with the suite agents and the agent builder. A lot of what you do with workflow today, where you've got this really sort of user intuitive, you know, intuitive interface for building, you know, non AI based workflows that's going to come to sort of AI workflows as well, where you don't now have to go in like right now it's really great.
[00:18:32] Guest 1: Ranga Bodla: And you go in and you tell it what to do. And it does a lot of that. And then it goes and it runs with it, and you're not even going to have to tell it what to do anymore, because you're going to be able to run these things continuously or on a schedule. You know, the prompts will be built into the workflows and then you're just it's about, you know, keeping a human in there to verify the outputs, etc.. So I think we are on the way there. I think right now, though, I mean, just at at having, you know, I love working with Claude in particular and starting at a high level. Here's what I want to do and then refining it. You know, you kind of gaslight the AI a little bit. You know, I treat my AI very unlike how I want to treat my partner in life or my friends. Right. You really want to manipulate it, right? You really come down on it when it's wrong, and then you really build it up when it's right.
[00:19:20] Guest 1: Ranga Bodla: And you're constantly optimizing the context and the instructions, the prompts as you go through that. And next thing you know, you've got these projects that are just coming up with consistent results, accurate and then dynamic, right. Drilling down to the underlying data in NetSuite opening, you know, showing me, you know, I've got a table of my top vendors by spend. And then it shows me exactly how it calculates it. It lets me drill into the transactions that it used to calculate it. It gives me all the anomalies, lets me drill into those anomaly transactions. I mean, it's gotten pretty incredible to what you can do. And we're just starting. I mean, where we're going to be in a year from now, two years from now, I think and I think with NetSuite having this really sophisticated data model, having such an open platform, you know, our DNA has always been the extensibility of the suite. And I think, again, AI now the potential of what you can do now to extend the capabilities of the suite with AI, it's just we're not even we're just at the beginning and it's really exciting.
[00:20:18] Co-host: Glenn Hopper : Yeah. And I think that's important to realize where we are in the evolution of AI and tying it into your workflows, because early on, we figured out ChatGPT Gemini are pretty good at if you if you load in some financial statements, it can look at the trends. It can do ratios for you at the financial statement level. It was giving you some good analysis, but we didn't have the context and the ability to drill down. And you didn't have the context window that's big enough. You're not going to put 10,000 transactions G-L transactions a month in there or whatever. So being able to tie directly, and I think especially with the MCP connector, being able to sort of have that schema to go across your database and all your tables know where to get the source information, then it can tell you your cogs are off. But here's what drove it. And it can look not just at this month, but at historical and say, hey, you've increased your spend with this one vendor by, you know, 60% this month. That's the big, you know, in Q and those things. And that's where it becomes a valuable partner. And that's where that's where everything's going. And it's, you know, you guys hear the same news that I do about all these AI projects that are failing right now. And I just I get frustrated because I understand 100% they are failing in a lot of ways, but they're not failing, in my opinion, based on where the AI technology is, it's people misapplying it or trying to just sprinkle some AI on something without engineering in it and all that. And that's I've said from the beginning, it's going to be the SaaS providers. For 98% of the businesses out there. The SaaS providers are the ones who are going to bring that AI to them. So there's been we're finally getting to that point now where you're starting to see it really baked in and adding that difference.
Ever feel like you go to market teams and finance speak different languages? This misalignment is a breeding ground for failure in pairing the predictive power of forecasts and delaying decisions that drive efficient growth. It's not for lack of trying, but getting all the data in one place doesn't mean you've gotten everyone on the same page. Meet QFlow.ai, the strategic finance platform purpose-built to solve the toughest part of planning and analysis of B2B revenue. Q flow quickly integrates key data from your go-to-market stack and accounting platform, then handles all the data prep and normalization. Under the hood, it automatically assembles your go-to-market stats, makes segmented scenario planning a breeze, and closes the planning loop. Create air-tight alignment, improve decision latency, and ensure accountability across the team.
[00:23:08] Guest 2: Joe Friedman: Well, I just want to add one other. I want to actually come back to a point you made earlier, Joe, which is about the data. And you know, the data is such a critical part of this. You know, that's the other thing that I think some of these projects have failed is, frankly, just you. You stick AI on top of bad data and you're just going to get bad results. So I think that is also one of the other elements here is, you know, NetSuite provides that, really that rocket fuel for AI. You know, we often say, you know, the best AI comes from the best data, and the best data comes from NetSuite. And we really believe that. And what we it's what we see. It's you know, I talked to your friend Jared Backlund. You know, Glenn and you know, that was one of the conversations he talked abou,t is he's like, yeah, I got some other data within, you know, I got to do all this cleanup work, but at least my NetSuite data, I know it's clean. I can just leverage that and, you know, do my work on top of that. And I think that's a really important point, is that that's really a core part of the data that you're going to leverage for AI and what you do. And, you know, I do want to come back. One comment, Glenn, you know, you were talking about Satya Nadella and and, you know, I mean, obviously it's it's an easy thing for him to say that SaaS is going to be dead if they're not really investing in their own SaaS applications. So, you know, it's sort of a, you know, an interesting commentary when they're not really investing in applications like we are. But, you know, I don't, I don't you're not going to vibe code ERP. That's I think, you know, so if anyone believes you're going to vibe code ERP, go ahead.
[00:24:33] Host: Paul Barnhurst: I was halfway done. I'm just kidding.
[00:24:37] Guest 2: Joe Friedman: You'll be done by the end of this podcast. Yeah. No, you know, I just I don't think you're gonna vibe Cody RP. I think the reality is you need that. You need the core processes that are there and leveraging AI, you know, in these processes, I think is going to be critical. It's just going to continue again, where we see value from automation. We'll see that value from AI in the same way.
[00:24:57] Guest 1: Ranga Bodla: Yeah. And I think, you know, we talk about ROI in these build AI projects. And I think maybe some of those were a little early. I think right now you can get incredible ROI from AI projects. And it's relatively low risk, right. If you're already a NetSuite customer, you download our AI connector for free. You hook it up to an enterprise cloud account, which is 20 bucks a month or whatever per user, right? And you don't actually have to change any of your underlying business processes at all. Right? You can point cloud at a standard report ID, at a custom report ID at a search ID, right. You can run Swql, you can run search. And you don't have to spend any money on it at all outside of what you're already spending for NetSuite and what you're spending for cloud, and you can literally make magic happen. I know that because I see it every day. And then on top of that, we're going to bring that experience directly in NetSuite at no further cost to the user right by, you know, and I know, you know, some of the new players like to say that, you know, NetSuite has been sitting on their laurels or or, you know, we got lost. And that couldn't be further from the truth, really.
[00:26:04] Guest 1: Ranga Bodla: We have been focused on a long term, scalable way to really bring AI to NetSuite for many, many years. We did it from the ground up, from the studs up, from the infrastructure layer up, leveraging this billions of dollars of investment and Oracle technology so that you don't have to make these huge investments in these failed AI projects, you're going to be able to get huge returns just from using NetSuite, just from using NetSuite, maybe in some of these industry leading tools for simply the cost of the licenses of those tools. And I think we've just gotten to the point where that's really possible. And so I do think some people maybe bought into the hype a little early. And that's good, right? There's nothing wrong with that. I think we're at the point now where the models are good enough. Right. The context are able to give to these models has reached a new level. And your ability again to connect them now via technology like MCP directly into the source of the data, has made it. So you can start to do these projects at virtually no extra cost outside of just the licenses for cloud and the licenses for NetSuite. And so I think the ROI is just going to start to skyrocket.
[00:27:14] Host: Paul Barnhurst: I'm curious kind of speaking to that. For the average customer out there, where do you recommend they start? Right. Is there seeing kind of the agent come out all these things? Is it just hooking it up to cloud and experimenting, you know, MCP server, what's the advice you'd give? You know they're on NetSuite. They got their data. Like, I know I need to take advantage of AI. And how do you generally kind of advise them to get started.
[00:27:39] Guest 1: Ranga Bodla: If it were me right and I was a relative, I was just like a smart accountant. But maybe he didn't know about AI, right? And I hadn't had sweets. I would, you know, I'd first internally we'd agree on, hey, let's sign an enterprise agreement with Claude, right? Or ChatGPT. Right. We support ChatGPT. We're going to support all of the models out of the box, and you could connect, you know, you could build your own MCP servers, etc., but out of the box right now, the AI connector is free. It works extraordinarily well with Claude. And it's we're getting there with ChatGPT. But I would say get an enterprise agreement with Claude. Right. Download the connector for free. You can set it up in less than, you know, less than 45 minutes. And I think the most bang for your buck with just the is start analyzing reports. Right. I would start with, you know, I'd open the report, right? I'd look at the little report ID and I would start with, hey, Claude, I'd like to use you as my financial statement, my financial expert. Financial analyst. Right. And I would like deep insights into my financials this period. Help me. What are some things that you would suggest for me to ask about and start helping me create a prompt, right, to get some of these deep analysis in my bed. Here are some of the things we look at today, right? Maybe even load in. You know what your outputs are today and what some of the reports are today. And it's going to be like let's go. Right. It's going to pat you on the back.
[00:29:04] Guest 1: Ranga Bodla: It's going to get you hyped up and it's going to say, let's do it, and it's going to come back and it May the 1st time may not give you the most perfect results, but then you say, hey, I like what you did here. I don't love what you did here. Some of these results were inaccurate. Do you understand why it's going to be sorry? Yeah, I do know why I did that. Right. And it's going to be. Let me fix that for you. And then you say, okay, whatever you just did. I want to create universal project instructions in Claude. And every change we make, I want you to keep updating these things. And then eventually, what you're going to do is you're going to have beautiful outputs, right? You could even do things like take a screw. You put your website in there and say, hey, I want to I want you to give me branding guidelines that make this look like my website or, or, you know, and it's going to be great. Let's do it. You know, and it's you don't have to change anything in your NetSuite account. You just point cloud at the report IDs. They're right there in the URL and you start working. And I think that was the biggest change from our first version of the connector to the second version of the connector is because being able to read the standard and custom reports in NetSuite, where you know the information in those reports has already gone through workflow approval, already has gone through a certain level of diligence and cleanliness, and then giving Claude's ability to access that directly and which, at an incredibly accurate level, I think has really changed the game as far as just getting started.
[00:30:25] Guest 1: Ranga Bodla: And then you go from there. And what I like to do is I like to create like a, you know, take a report in NetSuite, configure it a little bit, and then I point Claude at my configured report and I say, hey, Claude, I want you to look at this report whenever we do accruals. And I want you to accrual any POA in this report. You know pending billing right. Boom. It's going to create a journal entry for you. It's going to give you audit trails. It's going to link. You can say hey I want to link I want you to link to these underlying journal. Great. It's going to give you instructions exactly how to do that. And you work with the system to get the results you want. I am not an engineer. I'm an accountant. I have no engineering background. I have, you know, don't want to admit that I haven't taken a lot of classes on AI, but what I've realized is how when you converse with the AI, you tell it when it's wrong, you pat it on the back when it's right, like it will work with you and it will help you. You know, it might take a little while, but you'll get better at it. And then you just upload the project instructions, update them, optimize your prompts, and it's just going to start working really well for you.
[00:31:24] Guest 2: Joe Friedman: And just one other thing I want to just add to what Joe said, I think is really important too, is when you're setting that up with Claude, you can tell it, hey, I don't want it to train the models. I just want just, you know, I don't want it to my data to be be used by other models, things like that. You can do those simple things. You're setting up the rules of what the control is, what access it has. So it's not just like I'm suddenly given the keys to the kingdom. You decide as a company, as a user, what you want to do there. And I think that's a really important piece, is all the stuff that Joe was saying, which is amazing what you can do, but you get to you don't. You're not just you don't have to like, give up the keys to the kingdom in order to get those results that Joe's talking about. You can set it up so that you know what, I'm going to do it in this very specific way with this set of users, this set of permissions. And, you know, I'm going to really lock this down and that's great. You should do that. And I think it's a great way to start and, you know, continue to play with it. We're seeing customers do that. And that's exactly how they're thinking about and trying to, you know, build what they do.
[00:32:24] Guest 1: Ranga Bodla: Yeah. Just to piggyback on that, when you set it up right, you give it it's got a certain level of functions that it can do create records, update records, read reports, all like, you know, standard reports, all reports. Searches. Sweet. Ql right. So you have this sort of universal set of tools and functions you give it access to. But then that's limited by the roles and permissions assigned to the user that's using it. Right. So it can read it and create access records. But if you're, you know, an AP clerk, it's only going to be able to create and access your AP records or your AP reports. And if you're an administrator, you're going to have all the tools of the kingdom. But it is. There's a large list of universal functions and tools that can do, and then that's limited by the permissions and restrictions the user. But then beyond that, if you're more technical, you can create custom functions and custom tools as well and get it to access all sorts of custom records too. So, um, you know, and we're going to send you, you know, it's all OAuth security. When we send your data to cloud, it's, you know, it's the highest level of security, just like we're going to do when we send your data anywhere to any third party system.
[00:33:25] Co-host: Glenn Hopper : I've talked about this a lot where there's a rightfully so in a lot of instances if misused, you know, there's all this concern about data going in into large language models. There are. I've talked about this ad nauseum, so I'm not going to dive too deep into it here. But if OpenAI, if Google if anthropic, whatever company you're using, if they're soc2 type two compliant and they're using OAuth and you know, the hype around as long as you have it set, you know, don't train the model on my data. The buzz around the data security risks are overstated. There are compliance issues and, you know, certain things. You couldn't put anything on the cloud. But I think a lot of times that fear or concern about data is overinflated to represent sort of the fear of AI in general. And I think I've been saying this a couple of weeks now, just what I'm seeing with clients. Gartner hasn't announced it yet, but I feel like the marketplace is tipping over that peak of inflated expectations. And I think it's because of people not knowing how to deploy being a little early, but what is going to get us over that and move us to the what is it? The plateau of productivity on the other side is going to be that the SaaS companies are going to start getting it.
[00:34:35] Co-host: Glenn Hopper : So seeing you guys rolling this out now and just hearing what's coming here and what's happening in other SaaS tools right now and new platforms, you know, and, you know, going after the same area. But what I wonder is you and I'm going to put you guys on the spot because I get put on the spot on this all the time. So I'm going to see your thoughts in one day. You'll hear Karpathy talk about, you know, where AGI is, you know, a decade out or whatever. And then Elon Musk will come out ten minutes later and tweet, the next version of grok is going to be AGI. And so it's just there's a lot of conflicting information out there. But seeing what you guys are seeing now with NetSuite and looking across the ERP landscape, get out your crystal balls. Let's look five years ahead. And do you think I know we're talking a lot about MCP and kind of outside in, but do you think in the future it's more the AI is embedded inside the core system Or is it really going to be about this network of connected external agents that are kind of working around it? What do you think is the future around that?
[00:35:41] Guest 2: Joe Friedman: My personal opinion, I think it's going to be some combination of both. You know, one of the things I come back to all the time, you know, over the years, you know, the benefit that NetSuite has really driven for companies is the ability to scale versus just grow. And I look at growth as like you're just adding you're adding people. You're adding revenue. Scale is about adding revenue and adding profits without having to add, you know, lots of, you know, people at that same rate. And that's something that it's been a value proposition that NetSuite has been able to deliver. You know, now for better part of a decade for our customers, I think that's only going to continue, and it's going to get even better with AI, where they're going to continue to be able to scale and drive that, and some of it's going to be inside the business process within NetSuite. Some of it's going to be, like you said, some of those external agents starting to interface things like that. I think you're going to see both of those. And that's why I really think both are really important. And we're looking at both of those things. And, you know, and Evan brought this up at the beginning when he built NetSuite, is, you know, Evan wasn't an accountant. And he said that and he's like, I don't know everything that people need, which is why I'm going to make this extensible, why I'm going to make it open and allow people to build what they need. And I think that's really what's going to help NetSuite continue to survive and really thrive, you know, moving into the next five years or plus.
[00:36:59] Guest 1: Ranga Bodla: I agree with you. I think it's both right. I think you still need an ERP system as a system of record. I still think you need data integrity compliance. You need to be able you know there's still going to be internal controls, right? You still need audit trails. You still need the people to be accountable for the things that are done in a business. But I do think there is going to be, and I think you're going to have people that are going to want to interact with that system in the best ways possible, and they're going to have agents in that system. But I do think a large part of it will have that ERP system as part of a greater interconnected network, using sort of these industry leading models as a UI in a lot of ways to these systems. Right. But having a really a data model, right, that has data record for all aspects of the business, super configurable to the needs of your business. Having all that business logic in there, right. That then the AI can leverage to drive better outcomes, right. It's only going to make the agents more accurate and more powerful. So I think it's going to be both. I think you need to have both. I think that's why NetSuite is incredibly well positioned, because we do. We are taking that hybrid approach. We are committed to having this be an extremely open platform and making this data and tools accessible. But we are also been working for years, really to bring this new user experience to the NetSuite UI and to our users directly with great insights.
[00:38:20] Co-host: Glenn Hopper : And I'm really I'm looking forward to seeing NetSuite next and all this. All right. The show's gone very quickly. Now's the curveball part of the show. We had AI, we took your LinkedIn profile and your Bios amphetamine to chat. I use ChatGPT this week and we ask it to come up with quirky questions for you. So normally there's just there's two questions and there's a and Paul and I do it a little bit different way. But since there's two guests, one of you might get two questions and one of you might or we might split it up so that you each get one. But we do this through a random generator. And Paul, how about you introduce you let them know. And I don't know. You guys can decide amongst yourselves who's picking the first number.
[00:39:01] Guest 1: Ranga Bodla: But we are we having are we gonna get ChatGPT roast us right now because, uh, it's a great. But that would be roasting.
[00:39:08] Host: Paul Barnhurst: I don't think it will roast.
[00:39:10] Host: Paul Barnhurst: You, but the questions could be unique. You might have to really think for a minute.
[00:39:14] Host: Paul Barnhurst: Why don't we let, uh.
[00:39:17] Host: Paul Barnhurst: You know, I think we'll let Rogan just pick a number between 1 and 25.
[00:39:21] Guest 2: Joe Friedman: I'm gonna go with 13.
[00:39:23] Host: Paul Barnhurst: All right.
[00:39:24] Host: Paul Barnhurst: So looks like 13. Question is for Joe. So you're off the hook, Joe. You can thank him for that.
[00:39:31] Guest 2: Joe Friedman: Thank you.
[00:39:33] Host: Paul Barnhurst: What's the most human thing about working in AI? Remember, this was created by AI.
[00:39:42] Guest 1: Ranga Bodla: The human thing. I think I brought this up, but it's really its ability to be manipulated, right? Like, I think, you know, you can gaslight. I actually, I read a headline and read it on a prompting forum that I. But I took it to heart. You can gaslight the AI. You can come down hard on the AI and then you can build it back up. Right. To get it to give you better results. Now again, I don't recommend you treating your wife or your husband or your partner in the same way that I treat AI, right. Because I am incredibly manipulative to get it to do what I want it to do. But I do think that is a very human aspect to the AI and its ability to be manipulated to bend to your every whim.
[00:40:27] Host: Paul Barnhurst: Love it, love it, love it, you know, and.
[00:40:29] Host: Paul Barnhurst: Never thought we'd have manipulation, human and AI so much.
[00:40:33] Speaker6: I mean, gaslighting, not something.
[00:40:35] Host: Paul Barnhurst: I thought you'd do with. Ai.
[00:40:36] Speaker7: I saw that, I saw that, and I can't. I wish I could take.
[00:40:41] Guest 1: Ranga Bodla: Credit for it. It was a it was a it was a Reddit forum of somebody and I don't know who it was. You know, Reddit user 1874 AI or whatever it was. But I have taken that and I have run with it and I use it quite often.
[00:40:54] Co-host: Glenn Hopper : That's great. Okay, so I do a slightly different approach here. And I just since AI generated these, I have it pick a random one. So and these are already assigned to who's going to answer which. Let's hang on. Bear with me one second. Oh Ranga, this is for you. All right. Actually, I kind of. I kind of like this one. Which historical figure would you most want as a NetSuite beta tester? Huh?
[00:41:20] Guest 2: Joe Friedman: All right, all right. That that who. What.
[00:41:23] Guest 2: Joe Friedman: Nets. What historical figure would I want as a beta tester who I'm gonna have to go with Martin Luther King Jr. I mean, I think, you know, it would be interesting to see how, uh, how he might use NetSuite.
[00:41:37] Guest 2: Joe Friedman: You know, there's.
[00:41:38] Guest 2: Joe Friedman: Always like, you know, there's always, like, topics about, you know, is AI biased or those types of things. So I'd love to see what he would do with it, you know, and see how he'd use it. So that's gonna be my answer.
[00:41:48] Co-host: Glenn Hopper : Mine was going to be Hunter Thompson. I don't know why, I don't know, you might end up with a bullet riddled screen or something if it's a great answer though. And yeah, I don't know, we'll I'll send you the rest of these questions. Some of them are pretty good. It is funny every week. And I think, Paul, I think they've gotten better as the models have gotten better. The questions do seem to get a little better. I think it.
[00:42:08] Host: Paul Barnhurst: Also, you know, sometimes the prompts are a little better. I don't I don't use a canned prompt. I create it on the fly each week. So it's always just a little different. So you never quite know what you're going to get. And we switch up the models.
[00:42:21] Co-host: Glenn Hopper : That's the human in the loop. We're keeping it.
[00:42:24] Guest 1: Ranga Bodla: I get it, we get asked a lot for like, you know, what prompted you use? What prompted use? I really think that people I think there's so much benefit to sort of building your own prompts and working with the AI to build your prompts. Like, you know, there's a lot of classes you can take on prompt engineering, but and maybe I'm wrong here, but I do think that when you if you have a clear idea of what you're looking for and you're committed to working with the AI, you know, and having that conversation, that relationship with it, you know, it's you're going to really be able to dial it in and really optimize them and find out how to give it the right context, right, and the instructions, but also come up with the right prompts that can be universally used across other things. So yeah, I do I think I can't stress enough, like using it, getting in there, working with it to come up with prompts to really get the outcomes that you're looking for, I think is the best way to learn how to use it too. Great. I'm all for you with no canned prompts.
[00:43:19] Host: Paul Barnhurst: Yeah, we're not a huge. We've had plenty of conversations about those, so we won't let Glenn get loose on that on the soapbox again for another hour. You'll get his real opinion on canned prompts, but we're with you to say the least. So I think that's a great stopping point. Thank you both for joining us. It's a pleasure to have you. We've had a lot of fun and appreciate, you know, bringing NetSuite on and getting your guys' perspective on this space and how things are changing and how you're thinking about AI. So, Joel, Ranga, thanks so much for joining us.
[00:43:49] Guest 1: Ranga Bodla: Thank you, Paul, for having us. Thank you, Glen, as always. And it was fun.
[00:43:53] Guest 2: Joe Friedman: Yeah. Thanks, everybody. We really appreciate that. Thanks for having us.
[00:43:55] Host: Paul Barnhurst: All right. Thanks for listening to the Future Finance Show. And thanks to our sponsor, QFlow.ai. If you enjoyed this episode, please leave a rating and review on your podcast platform of choice, and may your robot overlords be with you.