In this Google Cloud AI Agents in Action Series interview, theCUBE Research’s John Furrier speaks with Avalara’s Marcus Larner about automating transactional tax and compliance. Larner explains how Avalara combines its deterministic tax engine, ALFA AI foundry and Avi agent to deliver real-time, audit-defensible accuracy across complex workflows at scale.
A live demonstration during the interview shows Avi working with Gemini Enterprise and Google Cloud, and Larner details Avalara’s active-active architecture, marketplace distribution and governance controls. The conversation shows how domain-specific data grounds large language models, helping tax teams improve speed, resilience, scale and trust without sacrificing precision or compliance standards.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
Google Cloud: AI Agents in Action. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Register for Google Cloud: AI Agents in Action
Please fill out the information below. You will receive an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for Google Cloud: AI Agents in Action.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
Google Cloud: AI Agents in Action. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Sign in to gain access to Google Cloud: AI Agents in Action
Please sign in with LinkedIn to continue to Google Cloud: AI Agents in Action. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Marcus Larner, Avalara
In this Google Cloud AI Agents in Action Series interview, theCUBE Research’s John Furrier speaks with Avalara’s Marcus Larner about automating transactional tax and compliance. Larner explains how Avalara combines its deterministic tax engine, ALFA AI foundry and Avi agent to deliver real-time, audit-defensible accuracy across complex workflows at scale.
A live demonstration during the interview shows Avi working with Gemini Enterprise and Google Cloud, and Larner details Avalara’s active-active architecture, marketplace distribution and governance controls. The conversation shows how domain-specific data grounds large language models, helping tax teams improve speed, resilience, scale and trust without sacrificing precision or compliance standards.
In this Google Cloud AI Agents in Action Series interview, theCUBE Research’s John Furrier speaks with Avalara’s Marcus Larner about automating transactional tax and compliance. Larner explains how Avalara combines its deterministic tax engine, ALFA AI foundry and Avi agent to deliver real-time, audit-defensible accuracy across complex workflows at scale.
A live demonstration during the interview shows Avi working with Gemini Enterprise and Google Cloud, and Larner details Avalara’s active-active architecture, marketplace distribution and governance c...Read more
exploreKeep Exploring
What does Avalara do, and how are you applying generative AI and autonomous agents in the enterprise while ensuring the accuracy, trust, and performance required for tax and compliance?add
What is the perspective on the opportunities and implications of agentic, domain-specific AI for customers and for the tax/finance/legal industry as a whole?add
How is AI changing the finance function—from simplifying workflows and shifting finance from reporting to strategic forecasting and revenue focus—and how has that affected technology and AI value creation, particularly with respect to real-time tax and regulatory integration?add
How are AI/ML expertise, public LLMs, and proprietary tax datasets combined to build precise, trusted, real-time tax solutions?add
>> Are your AI investments delivering insights or are they driving action? The conversation around AI is shifting from models and tools to outcomes and autonomy. We're moving beyond applications that just inform to intelligent agents that perform. I'm John Furrier, your host of theCUBE. Welcome to the AI Agents in Action series brought to you by Google Cloud, the show where we go beyond the hype and talk to technology leaders who are actually building autonomous agents to solve real world business problems. Please welcome Marcus Larner, who's the SVP of engineering at Avalara. Thanks for coming on Marcus, appreciate you. Agents in Action, very popular show. Thanks for coming on.
Marcus Larner
>> Yeah, hey John, pleasure to be here. Thanks for having us on.
John Furrier
>> So set the table on what you guys do, because there's a lot of things going on with generative AI and certainly agentic in the enterprise. Set the context and we'll get into some of the opportunities.
Marcus Larner
>> Yeah, so Avalara is a global leader in AI -powered transactional tax and compliance automation. So what does that mean? So when you think about transactional tax, it's sales tax, it's use tax, it's VAT tax, it's cross -border tariffs and duties. So all of that complexity is what Avalara simplifies and automates. We support directly or indirectly millions of companies throughout the world, across the globe, thousands of tax jurisdictions. And I think this is a particularly interesting place to talk about AI and agents, because when you talk about tax in any financial system, really, you have to be penny perfect. You have to have a high degree of trust. You have to be highly performant, highly resilient. And those are things you don't typically associate with AI. So AI can hallucinate. We know that. AI takes a little bit of time if you've got these deep LLMs. So we have to balance, where to apply AI smartly, intelligently, so that we maintain that high performance, that high degree of trust, and that high degree of accuracy. And with Google's partnership, that's what we're here to talk about and how we've achieved that.
John Furrier
>> one of the things I love about this market, LLMs large language models. That's like English and the spoken word. tax is its own language, the language of business. And it's complicated. There's a lot of manuals, a lot of data, a lot of facts. This kind of sets up the perfect kind of domain -specific value we're seeing in generative AI. And you're seeing really robust growth in areas like legal, finance, and tax. This is a really great opportunity. I'd love to get your point of view and perspective on the agentic AI opportunity for your customers and the industry as a whole. Domain -specific is the hot topic. That's where the value is being created.
Marcus Larner
>> Yeah, and this industry is changing rapidly. So if you look across the globe, you've got government organizations that are moving from what has historically been a cadence of compliance, whether companies file tax returns monthly or quarterly or annually in some cases. But where the industry is going is real time. And so governments want their data, their information and their tax revenue a lot quicker. And so you've got real time integrations. We call that e-invoicing. Governments are also lowering the threshold for compliance triggers, so that increases audit risk. So while this is happening, we have to become a lot more accurate and build and maintain that trust with companies. Now, within Avalara and within businesses, AI is providing that simplification. agents are increasingly able to take on complex workflows and free up the people and the human resources to do the complex tasks. And so we're simplifying workflows, we're making things more streamlined, and we have a changing role of finance. We're going to dig into that more as well. And finance now becomes a strategic player. It's not just a reporting mechanism, but how can you start to forecast and model and use all of the data and AI to strategize and mold your business.
John Furrier
>> I want to get your take on this because we're seeing the pressure with AI to drive revenue. In the old IT project, it was cost takeout. Revenue is a huge piece of it. So you've got the CFOs who care about their revenue, but also, the regulators, they want their revenue too, right? This is tax. Real-time, regulatory, and then CFOs play a much more active role. They're leaning in and saying, where's their cash? Where's the revenue? That becomes really big. Talk about how that's impacted some of the tech and the AI value creation.
Marcus Larner
>> Yeah, like I say, so doing that real-time integration for government organizations, that's a global trend. And giving data and information to CFOs in the moment, we all know that a lot of change has happened with duties, for example, and tariff modeling. And that's one example. But how you can optimize and factor in your tax burden and the different use cases and products that you have in your suite and use that as part of your strategy and your roadmap. That's where it's going. And that's where Avalara is playing in the middle of.
John Furrier
>> You take advantage of the agility needs as well. That's key. How are you guys providing value to customers? Give some use cases and some of the impact unlocked for customers. Because, again, real time and the changing landscape, you never know what's going to change. You get the speed. You get the agility. What is the unlock? What is the use case? What are you seeing the most?
Marcus Larner
>> Yeah, so there's two big areas. So our flagship product, AvaTax, we've been working with multiple clouds and GCP is in the middle of our strategy here to provide what we call an active -active solution. So having the ability for our customers to be deployed simultaneously across regions, across cloud providers. So if there's an issue at any point in time, we can within seconds fail customers over to a different region or a different location and keep that resiliency high. Now, what you have with agents in the background is they're monitoring that entire system. And so we have the ability to have agents participate in decision -making on where is the best location, where is the most resilient and highest -performance location for customers to be at. You layer on some of the technologies around ADK and A2A that Google is providing and the Google Cloud Marketplace, and now you start to have integrations that are simplified, onboarding that is simplified, and the whole end to end becomes a much more streamlined user -friendly platform.
John Furrier
>> Talk about the scale piece of it, because real time requires an AI native piece, not a bolt -on. Talk about that impact and what that changes for agility speed, because your AI native is core to be integrated in. Talk about this bolt -on approach versus say, leaning in as a native element.
Marcus Larner
>> Yeah. So at Avalara, we've been doing AI and machine learning for well over a decade. And so having that expertise and experience of knowing where to apply AI, where is the real-time capabilities you need to maintain, and also an enormous data set that you can layer on top of. If you take a public LLM and then you ground it and you layer on your own proprietary data set, and we have decades of data on how tax jurisdictions and tax rates and rules are changing and evolving in the language that is used. And so we've become very, very precise when you layer on top of and you expand and extend these public LLMs. And so that is how marrying our expertise and our data sets, marrying those with what's available and with these frontier models, along with how we go to market and build our solutions, balancing the real time and the high degree of trust. That's how we achieve this balance of trust and performance.
John Furrier
>> You know what's great about this is you've got a front end and back end approach here. you got Google on the back end, but you guys are going to be part of the transactions. With GenAI, there's opportunities to be involved in a lot of different touch points. You're seeing so many situations where you meet the customers where they are. That's my favorite phrase. Well, translating that to your business, there's a zillion Shopify stores out there. E-commerce is now democratized and horizontal all over the marketplace. How do you look at that? Because this is not yesterday's e -commerce transactional thinking. It's a whole other level with agents.
Marcus Larner
>> 100%. And so not only about deployment and physically being where our customers are, so we deployed globally, but then being standards-based and participating in these marketplaces. And the Gemini Marketplace is a great example. So participating in that marketplace, and Gemini and Google have a very high standard on what it means to be and what level of performance and accuracy and scale you need to be in order to participate. So participating in those marketplaces gives customers, our customers, confidence that we have those capabilities. And then the marketplace simplifies, lowers the barriers to entry, simplifies procurement, and that enables us to scale and reach more enterprises more quickly. So it's a win-win.
John Furrier
>> The AI marketplace, AI agent marketplace by Google extends off their other marketplace. How has that been able to help you guys get your solution out there from a reach standpoint?
Marcus Larner
>> So, the marketplace and then partnering with GCP and the Google team on co-selling. So you've got all of the layers. You've got a fantastic infrastructure layer with, the Google Cloud (GCP) You've got standards that layer on top of that with A2A and ADK that we leverage extensively. You've got a marketplace that lowers the barriers to entry. And then on top of that, you've got this co-selling, this go-to-market motion that we can partner with the Google team and so that we can reach customers together a lot more quickly and effectively.
John Furrier
>> I have to ask about security, obviously compliance, governance built in. Now you're seeing that with agent security. With Google, how's that playing out?
Marcus Larner
>> Security is a big topic and you've seen this evolution over the past few years. The early implementations of these standards really didn't have the security story nailed down. But now we do. So now having that infrastructure and the security controls all the way through your stack, so we have 100 % confidence that our solutions, even with AI and LLMs involved, are highly secure and maintain that trust. It's critically important.
John Furrier
>> Let's take a look at a demo of the agent in action. So let's take a look at this and then we'll talk about it. Avalara's agentic AI platform for global tax and compliance combines a deterministic tax engine with agentic AI that automates complex compliance workflows. Powered by ALFA, our AI foundry built on 50 plus leading models and backed by billions of real world transactions, Avalara agents observe, advise, and execute directly where you work, including in Gemini Enterprise. As part of this ecosystem, Avalara's Avi agent is available on Google Cloud Marketplace, letting external agents access our platform via the agent-to-agent protocol. Avi is your single entry point into Avalara's network. It uses natural language to orchestrate the entire compliance lifecycle, from calculations to filing and reporting. In this demo, we'll walk through a standard workflow to demonstrate the accuracy our customers rely on. First, we'll validate an address using San Francisco City Hall. Avi instantly routes that natural language request, validates the address, and returns precise latitude, longitude, and jurisdictional data. Next, let's find the right tax code for a pair of running shoes. Avi queries Avalara's tax engine, drawing from Avalara's comprehensive content library of over 87,000 regulatory sources across over 190 countries that are updated regularly. The engine deterministically pinpoints the exact tax codes. Notice how the engine immediately surfaces the highest confidence tax code first. With the code identified, we'll calculate tax for an $80 pair of running shoes shipping from Seattle to San Francisco. Avalara applies the latest rules and rates to calculate the exact tax, while Avi displays the real-time breakdown by jurisdiction. Finally, let's check where we have established nexus and tax liability. Avi instantly maps our nexus obligations across states and local jurisdictions. To wrap up, let's pull a transaction report. We'll summarize our total sales tax collected for June 2026. Avi generates the report, breaking down the collections by state and local jurisdiction. In just moments, Avalara combined agentic orchestration with our deterministic tax engine to deliver an accurate, complete, and audit-defensible workflow in Gemini Enterprise. That's automation you can see and accuracy you can trust. All right, that was awesome. Marcus, take us through what's going on behind the scenes because there's a lot of complexity. I know tax law from my small perch here in our business. I can imagine for a big company, any company that sells across borders, there's just the complexity and the rate of change. things are always changing and you've got to roll up all this reporting, this manual labor. What's going on in this demo? What's happened behind the scenes?
Marcus Larner
>> Yeah, so the thing to remember is this is an extremely complex space and it's becoming more complex. So governments across the globe want to have their piece of the pie. That's how we think about it. And so they are layering on complexity and slicing up the pie to be ever smaller pieces. And so we have to maintain that and continually simplify the solution for our customers. So what you're seeing here is a single agent. It's called Avi. It's live in the marketplace, as you just saw. And that single agent, you can interface and access any of the full suite of Avalara's products. And so the Avi agent in the front end is receiving requests. It is deciding, okay, which piece of the compliance puzzle needs to solve this? What data do I need to bring in? What source of truth and groundings do I need to apply in order to have a highly accurate solution? And then you can see how fast it is, how responsive it is. So this is where we have unique capabilities of bringing speed and reliability and trust to the fore and exposing that in this very, very simplified interface through the agents.
John Furrier
>> Well, speed and scale are the number one requirements of succeeding with AI and certainly cloud and AI coming together. No wonder you have so many customers. I can see there's real value there. I have to ask, obviously the AI agentic marketplace is one piece of the Google Cloud Marketplace. More broadly, how have you guys been able to grow your business and your go-to-market fast with Google Cloud Marketplace overall?
Marcus Larner
>> Oh, across the board. So I mentioned earlier our active-active strategy, which was building a high degree of resiliency. So we partnered with Google to apply and deploy our solutions globally and that enables us to be where our customers want us to be. So, for example, a great example is Shopify. So Shopify is on GCP. And when Shopify rolled out AvaTax, they needed us to be exactly in the same physical location, physical data centers that they were in, in order to have that high performance and high resiliency. And so we were able to support that. And we can do the same thing for customers who need that level of integration. And it's truly, it used to be the difference between an on -prem solution and a cloud solution used to be this discussion of uptime and availability, but no longer, not with Avalara, because we can be exactly where customers are, right next to them in their physical location, and be cloud enabled. And GCP has enabled us to do that.
John Furrier
>> It's such a win-win because any Google customer is your customer, vice versa, they win, you win. Great efficiency, the deployment speed's phenomenal. Again, it takes away a lot of that friction. That's a huge piece.
Marcus Larner
>> Exactly right.
John Furrier
>> All right, let's wrap up because I think this is going to continue to be hot. You're going to have money involved, compliance. You have complexity, you're taming that, you're taking the friction out. What's on your mind these days as a leader in the engineering team, you got to create the agents at scale, that's coming fast and they don't sleep. We go to bed, and the world operates and you get the cloud scale. So, what is your vision going forward? How are you guys going to keep innovating?
Marcus Larner
>> All the way through the stack. So you start with how our users and our customers experience the product and use the product, and how can we simplify their lives? That's the number one thing. How do we bring all of this together? So an end -to -end solution, no matter what they're doing, they don't have to worry, our customers don't have to worry about the tax compliance business. business, we handle that for them in the simplest way that we can. We free up their time and their energy to spend on what they need to, their core value proposition and what they need to innovate with. Internally, we continue our innovation. We look for ways where we look at where the public language models are going, the frontier models. We have, like I said before, enormous data sets that we can add to those things and build trust and make sure that the systems are continually grounded. But then also as we operationalize and build and manage our solutions, how can agents help make sure that we have a bulletproof system that's always up, always performant, and always available? So it's multifaceted across the whole stack.
John Furrier
>> Well, Marcus, this is a great demo. You guys are definitely an agent in action company. I love the tagline on your background for the company, build it, run it, and scale it. I think that's the North star of every company right now who is operationalizing and executing with AI. And certainly the agents are coming and they're building out fast and trust, quality data, reliability. You got to get the numbers right, revenues on the line, cash, money, and efficiency. Thanks so much for taking the time. Really appreciate it.
Marcus Larner
>> Wonderful, thanks John.
John Furrier
>> Okay, Marcus Larner is here, SVP of engineering at Avalara, solving a big problem. And again, this highlights the domain-specific, but you get the horizontal scale of cloud and AI and the domain expertise all coming into agents super fast. And this is the future Agents in Action. I'm John Furrier, your host of theCUBE. Thanks for watching.
John Furrier