How Modern Finance Teams Are Automating Billing and Revenue Workflows with AI Tools - Riya Grover
In this episode of Future Finance, hosts Paul Barnhurst and Glenn Hopper sit down with Riya Grover to explore how AI is transforming order-to-cash workflows. The conversation explores billing automation, revenue operations, and the evolving role of finance teams. Riya shares her entrepreneurial journey and how Sequence is building the first agentic AR platform. This episode is packed with practical insights for modern finance leaders navigating AI adoption.
Riya Grover is the Co-founder and CEO at Sequence. Sequence, backed by a16z with $40M raised, is building the first agentic platform for accounts receivable. The company helps B2B businesses automate quoting, billing, invoicing, and revenue recognition, especially for complex pricing and custom contracts. Prior to Sequence, Riya founded Feedr, a venture-backed company that exited to Compass Group in 2020. She holds an MBA from Harvard Business School and a BA in Economics and Management from Oxford University.
In this episode, you will discover:
Why building a two-sided marketplace is incredibly difficult
How modern B2B pricing models break traditional billing systems
The difference between deterministic systems and generative AI in finance
Why human-in-the-loop design is critical for financial AI agents
What the future finance tech stack will look like in the next five years
Riya explains how complex contracts, usage-based pricing, and custom deal structures create massive billing challenges for growing companies. Sequence solves this by combining deterministic billing foundations with AI-powered workflow agents. The discussion highlights where AI should and should not be used in finance operations. Trust, auditability, and human oversight remain central to successful AI implementation.
Follow Riya:
LinkedIn: https://www.linkedin.com/in/riya-grover-a22a4822/
Website: https://www.sequencehq.com/
Sequence Series A Fundraising: Announcement
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:
[02:05] – Riya’s Startup Journey
[06:55] – Marketplace Challenges
[08:28] – Modern Billing Gaps
[10:46] – AI in Finance
[13:10] – Human-in-the-Loop Design
[18:35] – ML vs. Gen AI
[24:26] – Series A & Scaling
[30:11] – Future Finance Stack
[35:15] – Rapid Fire & Wrap-Up
Full Show Transcript:
Host: Paul Barnhurst (00:00): Welcome to another episode of Future Finance. I am one of your hosts, Paul Barnhurst, and I have here with me my co-host, Glenn Hopper. Glenn, how are you doing?
Co-Host: Glenn Hopper (00:51):
Good, Paul. How are you? Happy Monday.
Host: Paul Barnhurst (00:53):
Yes, happy Monday. Excited to get an opportunity to chat with and introduce our guests. So start by doing a brief introductory shared background, but Riya Grover, welcome to the show.
Guest: Riya Grover (01:05):
Thank you. Great to be here with you both.
Host: Paul Barnhurst (01:07):
Excited to have you. So let me give a little bit about her background. She is the co-founder and CEO of Sequence. Sequence is backed by 16 Z with 40 million rays to date and is building the first agent platform for accounts receivable, helping B2B companies with custom pricing and sales contracts to automate, quoting, billing, invoicing and revenue recognition, all those things we love. Right Glenn?
Co-Host: Glenn Hopper (01:34):
Yeah, absolutely.
Host: Paul Barnhurst (01:36):
Prior to building Sequence r, built Feedr, a venture-backed software company that exited to public company Compass Group in 2020. Riya holds an MBA from Harvard Business School in 2015 and a BA in Economics and Management from Oxford University. So what that tells me is he's really smart. Alrighty, well again, welcome to the show. We're really excited to have you here.
Guest: Riya Grover (02:02):
Yeah, likewise. Just never tee it up by saying you're really smart; you set the bar very high there.
Host: Paul Barnhurst (02:08):
Well, I mean you have Harvard and Oxford, those are schools, and Glenn, you have Wharton on yours, is that the one or is it Harvard? Harvard. See I feel like I'm in a room with really smart people so I mention it. Alright, so enough of that, we'll jump into the first question we have for you. So you mentioned in your, your first company was a feedr and that was a technology platform in the restaurant space that you sold in 2020 to Compass Group. Maybe could you talk a little bit about why you started Feedr and what you learned from that first venture? Let's start there.
Guest: Riya Grover (02:42):
So I very simply saw a problem. You had lots of independent restaurants which had hours of downtime in their kitchens and they were trying to tap into new sources of demand. Often restaurant teams don't have large administrative functions where they can kind of go out and find those new opportunities. So we started looking at ways in which we could help them grow revenue by tapping into new segments of the market, so corporate and venue demand, and essentially built that two-sided marketplace and then also kind of built the software stack to enable them to services. So manage orders, payments, loyalty, inventory by the end of it as well. And so it was really interesting building a two-sided marketplace because you have to be able to scale both sides in parallel. Also interesting building a business with a logistics and human component because there's no amount of software or technology that you can build that can prevent humans making errors and you have to build the buffers into your product to be able to account for that.
(03:49):
But we didn't raise that much capital. We were running fairly lean as a business and I think really what I learned during that was how to build a business, how to operate a company, what matters. I developed some very strong points of view on talent, the importance of really great talent early and really kind of creating this idea of talent density and hiring at an exceptionally high performance bar because I could see in the cases where we had that kind of exceptional talent, what a dramatic impact it made on the trajectory of the business. I think you also learn whether it's something that you want to do, do you want to be an entrepreneur. There are as many low points and hard times as great ones. And by the way, the answer to that was yes, because I went and did it again.
Host: Paul Barnhurst (04:37):
I would sure hope so if you did it again. Yeah,
Guest: Riya Grover (04:40):
And other than that, I mean when you're solving in a new problem space at a different time in a different market, most things you are doing are from first principles but really that talent piece and also just really understanding what it takes to scale a company and ultimately exit a company is something that I got from that first experience.
Host: Paul Barnhurst (04:59):
I just wanted to ask a follow-up question because there's something, you mentioned the two-sided marketplace first I'm curious, have you ever read the book, the Cold Start Problem by Andrew Chin? Yeah,
Guest: Riya Grover (05:07):
I have.
Host: Paul Barnhurst (05:08):
Okay. That's what I was thinking of as you mentioned that and just would love your thoughts. For anyone who hasn't read it, I'll tee it up just a little bit. So Chen, he worked at Uber or no, he was at Lyft and several other marketplaces and works for one of the big VCs now helping people start marketplaces and he goes through and he talks about the challenges of starting a two-sided marketplace with one of the biggest being the cold start. You have to start from nothing to how you get to the most important side of the marketplace, how do you make sure you get the other side and skill it. So I would love to get your thoughts that book in relation to just starting a marketplace and the biggest challenges there, because I've always been interested by marketplaces,
Guest: Riya Grover (05:53):
It is really hard in the vein, no side is interested if you don't have critical mass on the other side, and that's the reality. So you're trying to scale those in parallel and give enough to both sides for it to be meaningful. It's probably not valuable until you have a critical mass on both sides. And for us we started by onboarding and scaling with the restaurant side first, but then of course, very quickly, you get the pushback of well,there's not much demand coming in. Is this worth my while? And so that's when you're sort of rapidly trying to build the other side of the marketplace in parallel. And what I will say is once you do have critical mass on both sides, there is this really beautiful viral loop that kind of kicks in where you are growing both sides quite naturally because you've kind of got that critical mass and I didn't start a marketplace business again, it's hard to do it really well, but it can be a really beautiful business model once if you can kind of drive to a certain level of success
Host: Paul Barnhurst (06:55):
Telling that the second business was something different. It's a very challenging business and can be very rewarding, but you have to solve multiple things at the same time to make it work. You can't just focus on one side or the other.
Guest: Riya Grover (07:05):
Absolutely.
Co-Host: Glenn Hopper (07:06):
And Paul, not for this episode, we have a lot to cover with them, but I've got to tell you the story of the business that was a marketplace. It was a user generated content and listeners company that I co-founded that we did not hit that inflexion point and we read the book as well and we had the chin model that we use to forecast all of our growth that came. We never hit that inflexion point, and now every time I hear Chin, I just think about that chin model and how just wrong we were on it. But that's a whole story for another day. Building a marketplace was very tough, getting the creators in our case and the listeners in as well, it was a lot. So kudos, kudos to you R for pulling it off in the restaurant space, but I love having that success and having that exit and then after that happens, I'm sure you've got the bug to do it again.
(07:56):
Which brings us to the sequence that you've founded in 2022 and it feels like with what you're doing you're kind of riding the wave at exactly the right time. So sequence AI-powered billing and quote to cash platform. And I'm wondering, we know about the marketplace side of it, but what led you to this industry and also because a lot of competitors there, so you've got to find a way, but I think maybe the timing might be an advantage there, but how do you find a way to stand out in that crowded field of competitors?
Guest: Riya Grover (08:28):
Yeah, so I mean when we started the business we had this observation that the way companies sell today, whether it's SaaS companies, AI companies, FinTech companies, and obviously many other verticals, the way companies sell is not simple in advance subscriptions like most of the traditional subscription management companies or even what ERP support is generally geared around, if your pricing is very simple and fits into a box, you can manage billing, subscription management, invoicing in those platforms. The reality is most B2B companies end up working around that because they've got a three-month trial or an opt-out or multi-year pricing step up or some kind of variable pricing model with seats or usage, and all of a sudden, actually, that workflow of saying taking a contract that's signed to actually collecting revenue against it becomes this very manual thing. And so when we came into the market, this was sort of pre LLM era, we started building this product in 2022, but we were seeing this huge tailwind where companies were closing contracts with all sorts of terms and different pricing structures and didn't want to be constrained in their back office and their billing engine by what their sales team were doing, sorry, didn't want to be constrained by their back office as to how agile they wanted to be in their go-to-market.
(09:51):
They wanted to be able to close deals and bill for them accurately in an automated way. And so there's been a tonne of evolution in how software companies are pricing over the last few years and we stepped into the market at a time where we built a level of pricing and billing flexibility that really most legacy players can't handle and that was our wedge into the space and what's been really exciting for us is over the last couple of years as we have seen these big leaps forward in gen AI and essentially what LLMs can do that plays really well to the types of workflows that we're supporting and we make a pretty key distinction here. There are certain things that need to be deterministic, so billing calculations, invoicing calculations, that's still code and that is not LLMs taking contracts and trying to figure out what an invoice should look like.
(10:46):
We are able to do that deterministically because we built these very robust pricing and billing foundations. But where LLMs are really powerful is kind of aiding workflow and getting some of the jobs that people are doing really kind of down to minutes or seconds and I'm sure through this call we'll share examples of those, but today we have an agent that will take a new sales contracts that sign and set up that subscription automatically with all of those terms taken out. So a human operator really has to review it for 30 seconds and can start the subscription live, LLMs drafting follow-ups to unpaid invoices or answering billing queries, LLM agents doing a review of all invoices before they go out the door so that humans only have to dive deep and double-click into 10% of them. That's really powerful in the workflow because you are going from billing runs, invoicing runs, it would take five or six days post month, then to now can be done in minutes and sent out the same day. And so for us, the intersection of agents, the capabilities of LLMs but with a very robust foundation and sort of system of record, is a very powerful combination.
Co-Host: Glenn Hopper (12:01):
Paul, I'm trying not to step all over your next question here. So I'm going to bite my tongue and let you go ahead and go because I'm jumping at the bit to jump in there, but I'm going to let you go here.
Host: Paul Barnhurst (12:12):
All good. Glenn, something you said that I kind of lost what we'll get to next question, but I really like how you mentioned the whole deterministic because so often you hear people right now thinking you could throw gen AI at everything and it's like okay, remember it's probabilistic or you shouldn't use gen AI at all because it's probabilistic and I think we all know the reality is there's a balance, there are use cases where gen AI is fabulous and there are use cases you probably shouldn't touch with gen ai and I know as AI continues to grow, we've all seen how rapidly it's changing Claude releasing its Excel app and all these different things from all the different tools. How do you ensure there's always human oversight in the process with your agents? How do you make sure you find that right balance right? As we continue to see agents do more and more and they're getting better but they still have these challenges, whether it's hallucinations or the probabilistic nature, whatever it might be, how do you balance that?
Guest: Riya Grover (13:10):
Yeah, I mean I think humans in the loop are so critical in finance operations, it's actually why domain-specific platforms have a right to win versus maybe building agents directly in a horizontal agent builder. It's because the work those agents are doing is so tightly coupled with the human workflow that they need to be designed and embedded in a way where you have really strong auditability, and you have this ability to actually get those points of human oversight and validation or judgment in things. So to be more specific, let's say our contract intake agent has taken a new contract and set up the billing subscription, humans still want to review that before invoices go out the door, even if you're driving for a hundred percent accuracy, that we're totally about contracts and revenue here. And so agents can do a great job at cutting out a lot of that manual work.
(14:07):
The agent can notify the human when a new contract is ready for review, but there is still a sort of human in the loop there with our AR agent, we have to think of this agent as someone working in your team. So they are following up with customers whose invoices are unpaid, they're replying to billing queries, following up with a purchase order number or a W-9 form if that's requested. So, kind of cutting out that busy work that otherwise a receivables team member would have to do when a request comes in, there are going to be scenarios where the agent can handle that with full confidence, in which case it's going to go back directly. There are many scenarios where nuance and input is needed so a customer hasn't paid and the contextual signals aren't there to understand exactly why. An agent will probably have a suggestion on how to follow up, but you may not actually want that agent to send that email out before it's been reviewed.
(15:04):
And so this is where I think product design really matters because an agent could pull in all the context, draft the message it suggests sending, give all the context to the human operator and just say, look, this is my suggested approach based on these inputs, are you happy for me to go ahead? And I think that's a really powerful paradigm of an agent pulling together all of that context, teeing up what needs to be sent out, but actually in certain cases not necessarily sending it automatically because the human has some potential context that lives outside of any data records that they might want to share. An account manager might have some context that they might want to share. And so this is where I think in pretty much every event capability that we've built into the platform in our designing the intersection with the work that people are doing and how they do it and really kind of embedding that very seamlessly into this, into the UI has been really, really important for us in sequence. And I think there are very few agents in finance operations where you write the agent and you automate it happily for it to be a black box and just run on the side without that sort of human oversight.
Host: Paul Barnhurst (16:16):
I think she just used the word that all finance people hate, right? Glenn Black box.
Co-Host: Glenn Hopper (16:23):
And it's really interesting hearing you talk about this from a product side because I think this might be a sign I've been doing consulting too long because every project I take on is a snowflake and it's somebody's precious baby and everything is so bespoke as we go through. But like you mentioned, there are things like if you define, okay, the AI agent can handle sending a W nine, it can handle these things. You don't have to reinvent the wheel every time you do it. Whereas when I come in with a client, we're building whatever it is a workflow or an automation or whatever and every time it's different. And instead of having sort of like you would with product rules around it, we're redefining for each client what this workflow does, where the gates are, what they're comfortable with. And I'm wondering if that's got to be hard from a product standpoint because you do have, you'll have customers with different risk tolerance around or trust tolerance around what they want to do. Are these things you toggle or is it
Guest: Riya Grover (17:23):
Yeah. When customers switch on an agent or a workflow, they will essentially configure that agent with their own business instructions and nuance and that can include tone of voice, that can include guardrails and things that they don't want to do. And we have some kind of pre-configured prompts to help them set out and define what they might want an agent to do versus not do. And this is where, as you said, tolerance for what the agent does when the agent acts versus when the agent suggests will vary between companies and you want the ability at a customer level to be able to configure it. That, and I think truthfully that is what is exciting about agents versus ML-based workflows or capabilities, which is that there is this kind of inherent reasoning ability and this ability to sort of more loosely prompt to get desired outcomes. But as you rightly said, it doesn't work in every scenario and there's a lot of parts of our product and our platform where we absolutely do not rely on gen AI and where gen AI wouldn't make sense for the kind of solution that we're trying to solve for.
Co-Host: Glenn Hopper (18:35):
I'd love to drill down on that a little bit more because right now better than most, I mean with classical ai, with machine learning, we've been using that scale for over well over to 15 years or more now using machine learning for everything from product recommendations to customer segmentation, whatever regression, what tools, the algorithms that we've used to great success that are deterministic on the machine learning side, but it was the barrier to entry to using machine learning was you had to be able to write Python and you had to have an understanding of the algorithms and data science and all that. Then generative AI came along where we were all sort of whether it was in our social media feed, what ads we were being served or what products are being recommended; we were all interacting with AI on a daily or hourly basis.
(19:26):
We just never thought about it. Now with generative ai, you can get in and you can talk to it like it's a person and you can even build your own, you can have it code your own algorithms and everything, but I think a lot of times people just lump it all together and they just think they try to apply the same rules that they do to generative AI to more of that classical machine learning. So how do you think about using both tools to get the most out of the technology, and obviously reducing just along with the black box, you've got the errors and hallucinations that generative AI is known for, but classical machine learning not so much
Host: Paul Barnhurst (19:59):
Ever feel like you go to market teams and finance and 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.
Guest: Riya Grover (21:07):
Yeah, I mean always kind of choosing the right technology for the problem that we're trying to solve and thinking about your point earlier when something should be deterministic versus probabilistic as an outcome. And so one example of that is transaction matching. So we have a reconciliation agent, a lot of that product will be designed and built with more classic machine learning where transaction matching and we're spotting for certain rules and patterns based on what bank payments have come in. But then gen AI can be quite helpful to sometimes provide judgement in some of the nuance cases and actually provide back information on why it wasn't able to transaction match or suggest an approach that a customer might take. We also have a sort of automations layer within the platform where people can build actual defined workflows and rules. So finance teams might want to put pricing guardrails on new quotes that are being signed to prevent certain discounting or to make sure that a certain kind of margin is achieved on a quote.
(22:18):
And so that's going to be rules-based, not going to necessarily be an LLM, they're prompting to share those approval workflows, other workflows that you'll have in the business, things that you're mapping to other systems of record. Again, sometimes actually building defined workflows is really great, exactly what scenario it's going to act in and exactly what the outcome is going to be. So we definitely leverage a mix of both and I think that's really important again, in the domain of financial operations where there's just no room for hallucination and error in certain places, particularly on things like invoicing calculations, usage calculations, filling, revenue recognition,
Co-Host: Glenn Hopper (23:00):
And when you find areas where you can stack 'em on top of each other and get an even more valuable response out of it, that's when you start feeling like I'm doing magic here.
Guest: Riya Grover (23:10):
Absolutely. Absolutely. And that is actually, I think, where the magic comes from, and that's why I think they're ultimately coming back to the product design, the ux, how and where agents and AI interject in these processes, how you get oversight into what they're doing. I think that's a really big part of winning here because this is all about trust. Back to that no black box thing, a big part of trust is understanding what AI and agents are doing or not doing and how you can interact with them in the system.
Host: Paul Barnhurst (23:46):
I agree. I think it's a huge part of it that comes back to that trust. So knowing when and how to use things goes a long way toward that. Exactly. So you recently raised 20 million, I think it's just the last few months here as part of a series A. So maybe talk a little bit about that. I believe if I remember correctly, you really didn't do the VC route with your first company, so this is probably one of the first big funding rounds you've done. I mean I imagine you did a pre-seed, but one of your largest, you've done so far. So maybe talk a little bit of what that experience was like, maybe some of the highlights and lowlights of raising those funds and then what's the plan for continuing to scale with those funds? How are you thinking about things?
Guest: Riya Grover (24:26):
Yeah, so we actually raised a lot to seed round as well. So this new $20 million investment brings us up to about $40 million raised in the business too.
Host: Paul Barnhurst (24:37):
Okay, so you did a large, I didn't realise it was that large of a seed,
Guest: Riya Grover (24:40):
But really continue investing in our product and engineering team to build out what is a very ambitious roadmap across the order-to-cash space and lots more revenue agents and enterprise-grade functionality coming later this year. And then also of course, invest in growing our go-to-market team. So sales, marketing and success.
Host: Paul Barnhurst (25:04):
And I'm curious, when you think of order to cash, how do you think of integrating with the big CRMs? What do you think about that? I know they have tools, lots of tools that integrate with them. You sometimes see about being separate from billing, so as you talk about the ambitious roadmap, how are you thinking about all that?
Guest: Riya Grover (25:22):
We've thought about a unified approach to CPQ and billing because the problem that we see in a lot of businesses is that sales teams live in CRMs, they're closing contracts upstream and that data is entirely disconnected from what finance are doing and seeing downstream and a lot of the revenue leakage, a lot of the manual work comes in actually just trying to connect and understand what sales have sold and making sure that it's represented accurately in financial systems of record. We really are playing a role in bridging that gap and giving companies the product catalogue, the pricing engine to be able to configure and track the terms of every contract sold, but then absolutely ensure that that data is pushed into the CRM is synchronous with what's pushed ultimately into the ERP so that actually rev ops and finance teams are working with this shared context layer and unified source of truth on really what product pricing usage and revenue data is.
(26:31):
And I think that's incredibly powerful. Actually, for an org where it's insane, you'll see these companies that are kind of a hundred million plus in revenue and actually that data lives in spreadsheets or PDF documents somewhere and has been unified, and you actually have no idea if you've billed for the right things or if you've missed some revenue somewhere. And what's held in these contracts is actually a really important signal for them how to drive further revenue growth in the business. And so I think one of the big value-adds we bring is not just enabling companies to track that data, but then actually pushing it back into the systems of record where companies are reporting and learning from that data to be able to actually leverage it in positive ways.
Host: Paul Barnhurst (27:16):
When you said you'd be surprised at what's done in spreadsheets, not at all supporting business, that complete disconnect from our CRM and our billing system and for the longest time we were pretty much billing it out all out of Excel files
Guest: Riya Grover (27:29):
And
Host: Paul Barnhurst (27:29):
Just uploading it to the system for the invoice, and at one point, it got so bad that I had to review all the files. There were so many mistakes. So I can definitely relate to what you're talking about in those nightmare billing scenarios. I've lived them
Co-Host: Glenn Hopper (27:44):
And I think it's such an amazing time right now because there are all, I mean Paul hears me say this all the time, we've been talking about digital transformation for 30 years and there've been tools along the way that have come along and made things easier, but I don't mean I don't want to get into the whole AI crash and all that. However, this is reminiscent to me in a lot of ways of the.com era with the rapid growth and so many people coming into the market. The difference is that it was kind of, well, we just have to be on the web. We don't know to what end, we don't know what it's going to mean. We just have to go to go.com, we have to do whatever we can. Whereas now the speed at which tools are coming out that facilitate these tasks and issues that we've dealt with for years, it's amazing.
(28:32):
And me on the consulting side, we're trying to build solutions, but at the same time, on the product side, I mean it seems whether it's an existing platform or a new startup, I've never seen, and I've been very tech-focused through my whole career. I've never seen a pace of such rapid technological change before. And even trying is a consumer trying to keep up with it and looking at the marketplace and all the software that's out there now and everything. I mean it's got to be invigorating as heading a startup in space right now, but just as a user of some of these tools trying to keep up with it, it's jarring. I bet from your side it's even more so
Guest: Riya Grover (29:12):
It's unbelievable and I think it actually really forces you as a CEO to force yourself to continue growing and learning at a rate that ultimately enables you to beat the market because things are moving so quickly. I think speaking on a panel recently where someone said as a CEO, you basically in this era have to reinvent yourself every few months. Things are changing so rapidly that if you don't keep pace with that market change, the new shiny things that you're building will become obsolete and all news very quickly. Also, just from a perspective of looking at the market and understanding your positioning in that market, again, that's changing faster than ever before historically, just with so many new players kind of trying to tackle this space in different ways. And so it's really fun being a CEO. You're on your toes for sure. But yeah, I'm enjoying it a lot.
Co-Host: Glenn Hopper (30:11):
And I guess as follow up to that, because you are forced to stay on top of what's going on, and so there's the software platforms, there's the stuff happening under the hood, there's what the frontier companies are doing, and this is a tough question, I know because the pace of change is so fast, but AI is taking on more of a role in finance departments. I think we're more risk averse than our sales and marketing companions, but now they are just being forced, sometimes dragged kicking and screaming into it. But now finance and accounting are adopting AI. I mean, what do you get out your crystal ball and what does the tech stack look like in five years? I mean like with cohorts in recent weeks, the perfect example of the software stocks plunge because oh, and it's like Satya Nadella said, well what's going to happen to SaaS, if gen AI is just, it's just a wrapper around a database basically. And I don't know, it's hard to fathom where it's going. But I also think, I don't want to answer for you, I've got my own thoughts around this, but I'd love to hear what you think about, not even necessarily what's going to happen to the SaaS industry or all that, but if this is what we're doing three years after chat g chatt 3.5,
Guest: Riya Grover (31:33):
Where
Co-Host: Glenn Hopper (31:34):
Are we? Another five.
Guest: Riya Grover (31:35):
Look for what it's worth. I do think that the model capabilities will plateau somewhere. I don't think we're, I think the rate of change that we've seen over the last few years, I don't think that necessarily sustains on the fundamental capabilities of the models and what they can deliver. But what is going to meaningfully change, in my opinion, is how companies and teams work over the next five years and it's just getting started. I think when I think about the future of the finance function, I think it is going to be leaner generally both in terms of software and team size. So speaking first on software you used to have, even in our space in order to cash, you used to have five, six plus tools that would enable you to basically manage revenue. You have a CPQ tool, billing tool, maybe a separate place you do rev rec, something else that you spun up to manage reconciliation somewhere else, you're calculating your sales commissions, et cetera, et cetera.
(32:35):
Whereas I do think that now you'll have fewer systems, maybe have one for revenue, one for spend, and you'll be able to do a lot more in those systems because fundamentally these businesses like mine today can build faster, more powerful capabilities in a shorter span of time and essentially build and execute on more product in a shorter space of time than you could 10 years ago. And so I do think we'll see less in the tech stack relative to your survey of some finance teams today and they're working with 15 plus tools. And then the other thing is teams and smaller leaner teams and where teams spend their time. So I think you'll start to see people move away from that sort of bottom of the pyramid work of this kind of repetitive manual operational stuff, answering billing queries and updating and editing an invoice with a credit node or manually literally ing in invoices and invoice data.
(33:36):
And in some cases that stuff will go away because agents will handle that really, really well. And you should start to see all this talk of the rise of the strategic finance leader. I think every finance leader will and should be spending their time on more strategic work. So in our case, less time on reviewing invoices and more time spent looking at pricing evolution and revenue tracking and understanding which product segments are driving your revenue to then inform pricing guardrails on what the revenue team are doing as an example of the difference in where time is being spent. Because I do think today finance teams are spending the majority of their time on things that are not particularly needle moving for the business. Just things that have to be done every month. These painful manual, repetitive tasks.
Co-Host: Glenn Hopper (34:29):
Yeah, you're preaching to the choir there, what a time to be alive. It's exciting. Yeah. I think in the interest of time, we need to get to our,
Host: Paul Barnhurst (34:37):
I know
Co-Host: Glenn Hopper (34:37):
AI questions, don't we?
Host: Paul Barnhurst (34:39):
How this section works is we pick a different AI model, we've used them all and we feed it the questions we're going to ask you, your LinkedIn profile, whatever can find on the web and ask it to come up with 25 kinds of personal, sometimes a little quirky questions. We get some odd ones from time to time. Usually they're pretty good. And so how it works is Glenn and I each ask a question, but we each take a different approach. So you got one of two options for the question I'm going to ask you. I can let the random number generator pick a number between one and 25, or I can have a human in the loop and you can pick the number between one and 25.
Guest: Riya Grover (35:15):
I'll pick the number.
Host: Paul Barnhurst (35:16):
All right, what number do you want?
Guest: Riya Grover (35:18):
Okay, I'll go with seven.
Host: Paul Barnhurst (35:19):
Alright, let's see what we got. So number seven's under the section, it labelled the foodie founder era. It says if you could create a food subscription box for stressed out CFOs, what would be in it? Okay.
Guest: Riya Grover (35:38):
I would have some good 80% dark chocolate in there, probably protein shake. It's efficient. We don't have any time for anything fresh, do we? Yeah,
Host: Paul Barnhurst (35:52):
That's your call. Yeah.
Guest: Riya Grover (35:54):
And maybe like a fruit. Keep it simple.
Host: Paul Barnhurst (35:56):
All right. I like a chocolate smoothie, protein shake, whatever you want to call it. Chocolate.
Guest: Riya Grover (36:01):
And the fruit should be fairly low GI because you got the dark chocolate there.
Host: Paul Barnhurst (36:06):
Nice. Alrighty. I'm going to turn it over to Glenn. He takes a little different problem with these questions, our different
Co-Host: Glenn Hopper (36:12):
Approach. Yeah. So first off I have to say, Paul, which model did you get to come up with these questions? These are way better than whatever chat GPT came up with.
Host: Paul Barnhurst (36:18):
It was Claude.
Co-Host: Glenn Hopper (36:19):
This was Claude. Yep. Claude wins again, I don't know.
Host: Paul Barnhurst (36:21):
And we asked something about if a golden retriever was doing accounting, what kind of golden retriever would it be or something?
Co-Host: Glenn Hopper (36:28):
Yeah, something just complete. Yeah,
Host: Paul Barnhurst (36:30):
They were just about that bad last time. So these are actually pretty good questions.
Co-Host: Glenn Hopper (36:34):
So I almost did an override to put you on the spot R I'll tell you the question. I'm not going to do this one, but the way that I do mine is I feel like, well, AI generated the questions, let it pick its favourite. But I was reading through the questions and the first one I felt like would've put you on the spot. It is. You've lived quite the academic journey between, and Paul, you left Yale, so you've had Oxford, Yale, and Harvard. If you had to survive on a deserted island with only students from one of those schools, which alumni group would you choose and why? But I'm going to skip that. I thought that would be a,
Host: Paul Barnhurst (37:04):
She's like, please don't make me answer.
Co-Host: Glenn Hopper (37:07):
So instead we'll let Claude pick this. Hang on. Oh, this is going back to the snack as well. Let's see. Okay, this is good. As someone who worked on improving the health and nutrition of PepsiCo's snack ranges during your MBA, what's your guilty pleasure snack that you'd never admit to a health conscious
Guest: Riya Grover (37:28):
Consumer? Honestly, it's the dark chocolate again.
Co-Host: Glenn Hopper (37:31):
I had a feeling that was coming and that, but 80% dark chocolate is a great one, right?
Guest: Riya Grover (37:36):
You can give one more. I feel like those are
Co-Host: Glenn Hopper (37:38):
Alright. These are all food related. Okay. The next one is, you've described loving the immense highs and lows of building companies. What's your go-to ritual when you're in the low's? Ice cream karaoke or spreadsheets. I don't know, that's kind of weird. Maybe I should add humans in the loop more on mine. Adam, do you have a go-to ritual to pull you out of the stressful parts? And these are all sort of related though, I don't know. Is that Yeah,
Guest: Riya Grover (38:02):
I mean I have two kids, so I feel like maybe not less stressful, but it's definitely a diversion or downtime from work. So that's where the rest of my mind goes. Two kids would
Co-Host: Glenn Hopper (38:15):
Do that. Well, these are all great questions, Paul. I think, I don't know, Claude has been just winning back to back challenges with me for weeks now, so I don't know. But these were good even though we did sort of in the random nature of it. Hit on a theme there, but that's okay.
Guest: Riya Grover (38:30):
Well, I guess everybody knows I like dark chocolate
Host: Paul Barnhurst (38:33):
Now. There we go. We know what you love and we know what you're building. So Maria, thank you so much for joining us and spending some time chatting about what sequence is up to what you're doing. Best of luck continuing to scale and grow the business. I know you have some ambitious goals for that, so we wish you the best of luck. Thank
Guest: Riya Grover (38:50):
You. It's been great to be with you both today.
Co-Host: Glenn Hopper (38:53):
Thanks Riya.
Host: Paul Barnhurst (38:54):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.
