Where Autonomous Meets Rules-Based: Why GTM Needs Both

Julia Nimchinski:
Thanks, and with that, we want to welcome our next session, Hilary Terrell, VP of Product Marketing, and Hugh Walton, Senior Solution Consulting. at Leen Data. Welcome to the show! How are you doing?
Hilary Terrell:
Hello, hello, we’re good, how are you?
Hugh Walton:
Good, thank you.
Julia Nimchinski:
Excited to dive into this!
Hugh Walton:
Yeah.
Julia Nimchinski:
So when autonomous… yeah, when Autonomous meets rule-based.
Hilary Terrell:
Yes, a little…
Julia Nimchinski:
Come in.
Hilary Terrell:
Spicy, a little spicy, contradictory, maybe. But let me hop. In, can I get a thumbs up that y’all can see? Slides, we’re good? Awesome. All right. So, thanks for joining us, excited to be here.
I think this is the second time LeanData’s been joining one of these, one of these sessions, which have been super fun, and obviously a hot, hot topic, for all of us in the go-to-market world.
So today, we’re going to talk about not just autonomous revenue agents, but where does autonomous need a rules-based component to really drive success in In this new world. So, I’m Hilary, I lead the product marketing organization here at LeanData. I’ll pass to Hugh in a second, because that’s my next slide, but here’s what we’ll be covering today.
We’ll talk through this point of view, why we believe both is required in today’s go-to-market world. We will jump right into a demo of showing how this shows up in action, and these are actual use cases that we use internally at LeanData, drinking our own champagne, and also use cases for… that we see with our customers. And potentially Q&A.
So, Hugh, my fellow person from Georgia, want to introduce yourself?
Hugh Walton:
Hi, thank you, Hilary, and thank you for inviting us, Julie, and thank you, everyone, for being here. I’m Hugh Walton, Senior Solutions Consultant here at LeanData.
I’ve been here with LeanData for over 5 years now, and we’ll be walking through 3 kind of mini demo use cases around specific personas and how AI and deterministic outcomes, you know, address, you know, the customer, you know, RevOps managers, and sellers. So, looking forward to that later on. Thank you.
Hilary Terrell:
And unplanned, we both wore our shirts today. We did not discuss that in our prep call. Alright, what’s kind of behind this point of view of the Autonomous requiring the rules-based?
This is what we’re hearing from all of the customers that we talk to, this is what we are experiencing internally, this is what we’re hearing, outside and in the market, outside of our own customer base.
Everyone is being really pushed at the board level all the way down to transform with AI, return to growth if that has stalled, and by the way, do it with the same, or shrinking budgets, or the same shrinking headcount, so doing more with less, but trying to go faster with AI.
So, we kind of see the world moving from the chapter, the first chapter of the promise of AI, unlimited potential, experiment everything, token maxing, go fast, move fast and break things. It’s like that phrase is, is new again.
to what we call the AI reckoning, and what we’re hearing from all of our customers today as they’re evaluating enterprise-grade solution is AI councils, and security reviews, and not just speed for speed’s sake, but speed that ties to business impact. And to do that, what are the operational guardrails that companies need to put in place?
So what does it mean when Autonomous doesn’t have the rules-based, kind of partnership behind the scenes? So a couple of examples, and we’ve seen them here ourselves.
Think about an automated marketing agent or marketing send that goes out to the wrong account, and potentially that account has a Tier 1 escalation issue, an at-risk account, but the agent didn’t have the right data to act on. Or think about an SDR AI agent on your website. We have one on our own website today.
But imagine if that AI SDR couldn’t tell if this was a customer or a prospect, and they’re trying to go through a qualification process, and someone’s getting frustrated, and maybe you had an upsell or an expansion opportunity. So all of these different missed signals, when the rules underneath it, are not powering that autonomous motion.
And we believe it all kind of bubbles into these three categories of where the autonomous agents can go wrong. It’s lacking customer context across what is marketing done, sales, post-sales and customer success, what is the business process, the rules that need to govern what the agent is doing, and the data foundation.
I was at the tail end of the outreach session, I could hear them definitely talking about people not trusting their data. This is common, no matter what solution. you were looking at, from an Agentic perspective.
So this is our point of view in terms of what companies need to move forward, as AI and autonomous agents are beginning to touch mission-critical revenue processes, from first, first touch, first signal, all the way to close one and expansion. We believe it requires a strong data foundation, so clean data.
For us, that means in your CRM, that agents and people can trust, everyone working on the same source of data truth. The second is business process truth. What is the end-to-end flow that we want a lead to go through, and how does that differ if it is a net new account versus an existing customer, or a North American account versus an EMEA account?
There are all sorts of different logic and workflows that often live inside people’s heads. So, how do we expect an agent to work, to work without having context for that business process? And then finally, a shared sense of what’s happened across the customer, whether it’s where marketing has engaged, sales, or CS.
So, both humans and agents having the same foundation from which to work from and move quickly. And we believe that’s what allows companies to move with both speed and confidence. So at the end of the day, we believe both rules-based and Agentic will coexist for some time, at least within our customer base.
They’re not entrusting everything in their go-to-market motion to agents. It’s a combination of picking what’s best for the job, where is the human in the loop, and where do we want to empower an Agentic workflow? And we believe LeanData is uniquely positioned to really help those two work together, and you’ll see that in the demo.
So for those not familiar with LeanData, that’s who we are. We’re an orchestration layer that’s really the thread across your entire go-to-market motion, even more important in this world of autonomous agents that are really touching revenue-critical processes.
So here is what we’re going to jump into in the demo, so I’ll pass it over to Hugh to give a quick introduction, and then I’ll stop sharing, and we can jump into the demo org.

Hugh Walton:
Yep, great. So, the first thing that we’re gonna do is we’re gonna look at a AI SDR to help with some automated scheduling. It’s gonna collect information. We’re gonna use that information in LeanData’s orchestration workflows to make sure that we get deterministic outcomes.
We’re scheduling with the right person based on what that AI agent has captured from your customer.
Second, we’re gonna look at, you know, how AI can assist a RevOps persona who, you know, now that things have been routed and things have been assigned and scheduled, I want to analyze that, and I want to see to make sure that my processes and my deterministic workflows are acting the way that they should be acting.
So I’m having AI analyze that logic and provide an output for me, an analyzation that tells me what’s working, what’s not. what could I be doing differently? For those deterministic outcomes. And last, we’ll go for a little demo around a seller’s experience, where they come into Salesforce, and they see a journey, right?
So it’s never just one person that is in a buying, motion anymore. It’s buying groups. We call them journeys here at LeanData.
So when a seller gets in, we’ll see that journey, we’ll see where people have, you know, responded to signals, have they gotten involved, what times they got involved in that journey, and we’ll use AI Assistant to kind of inform us, you know, based on other past one opportunities. What should be the next step?
Who are we not, including in that buying group? So… I will share my screen. Alright, thumbs up? Everybody can see. Good, good. So let’s start with Charlie here. This again, you know, I’m gonna interact with Charlie to, get us… get a demo scheduled, so let me speak with Charlie. Hi! How can I help you? Hey, tell me about LeanData, Charlie.
Absolutely. LeanData is all about intelligent GTM orchestration. We help teams automate lead routing, scheduling, and buying group coordination right inside Salesforce. Are you using Salesforce today? Absolutely am. Hey, Charlie, tell me some of LeanData’s larger customers. Sure thing.
A few of LeanData’s larger customers include companies like Anthropic, OpenAI, and Databricks. They rely on us for complex lead routing and GTM orchestration. Perfect. You know what? I think I’d like to book a live demo. Great! I can help set that up. What’s the best email to use? Please type it into the chat messenger for accuracy.
Thanks for that. Now, how many employees does your company have? You can pick from the options on your screen. Got it, Hugh. I’m checking for available meeting times now, so you might see a scheduler pop up. If no times are available, our team will be in touch as soon as possible.
Well, I did not see the screen pop up, so what we’re gonna do now is go to our Flow Builder. What’s happened, two things happened here, or should have happened. First of all. The AI agent took my information here, this Topic Lounge, and it used it to pass off to LeanData to say, who owns this account, Topic Lounge?
And what should have happened is a calendar should have popped up with the owner of that account. But, of course, it’s a demo, so it didn’t cooperate with me today. But let’s show you how that would happen behind the scenes. So, this is what you’re looking at here, is a flow builder within LeanData.
And what we have here are what we call trigger nodes. And what this one is, this AI SDR chat, is mapping all of that information that was collected by the SDR agent and putting it into variables that we’ll then use as data points to determine whose calendar I’m going to publish. For that person to schedule time with.
And I can use some of my, you know, logic tools to weed out bad or missing data. Maybe it was a bogus public domain email, so I just want to redirect them to a URL. I may want to look for duplicates, right? And if it’s a duplicate, I just schedule with that duplicate owner, right? Or, you know, what type of account is it?
Named account, customer account, and schedule the meeting with the right person. So again, we’re taking those… Using AI to gather the information, interact with the customer in a meaningful way, and then using those data points that we capture from AI to make deterministic decisions on scheduling, for instance.
The next thing I want to show is, a RevOps kind of, scenario where, you know. I might be in RevOps, and I might, you know… then that interaction happened, right? And I need to know why it got scheduled to Brennan Hawkins. Maybe there is… a question why, maybe there was a reason that it shouldn’t have gone to Brendan.
So, I have here my audit logs, and this is, you know, all real-time data that happens, and it shows you exactly when that lead comes in, the information that they gave. And all of the paths that it took through those nodes that you saw on that graph. So this is showing me what’s going on, but it’s a lot of information.
So I may want to use my AI agent as kind of an informer, right? I need to glean some information out of this routing process to help me better understand without having to crawl through all of the various information. So, I passed off, hey, how was the lead Steve Dallas routed? And it tells me exactly what it did.
The trigger, the lead was created, we went through a junk check, we did a deduplication, we found an account that it matched to, we went ahead and assigned it to that account owner, we sent out a notification, and we started SLA tracking.
We also passed it off to an AI inference node that looked at some of the responses and said, you know what, this person may be work… looking for support, rather than a sales motion.
So, you know, we can understand all of that using our AI assistant, analyzing those graphs and given us meaningful data that we can then act on, adjust to, etc, to make our processes even, even better, even more accurate. Absolutely. It streamlines the process, giving you clear insights without digging through all the data manually.
Yeah, thank you.
Hilary Terrell:
thoroughly.
Hugh Walton:
Where did he go? I gotta kill him. Anyway, so, let’s go to the, Let’s go to my next one. This is… Whenever you’re ready, just let me know what you’d like to dive into next, and we can walk through it together. Yep, sorry about that, everyone. This is our LeanData Journeys.
So, this is when a seller would log into you know, Salesforce, we’re in Salesforce, LeanData is a native Salesforce application, and I want to see what’s going on with my deal with smart technologies. That makes sense. Since LeanData is native to Salesforce, you can easily pull up that Smart Technologies deal right inside your Salesforce view.
You’ll see all the key routing, ownership, and next actions right there. That way, you’ve got full visibility into what’s happening with that deal. Actually, Charlie’s helping me out here, come to find out.
Hilary Terrell:
I mean, he’s writing the script as you’re saying it. Feel free to.
Hugh Walton:
Glad to be of help. Let me know what else you’d like to dive into or explore next. Sorry, guys. So let’s, let’s dig into this, particular, journey. So, I have here this journey when I click into it. It gives me the timeline of what’s been happening pre-opportunity, qualification stage.
It’s showing me all of the members that are part of this journey. Ai is telling me I’m missing a decision maker here. That’s important to this journey, to the success of it, so I may want to pay attention to that. It’s showing me all the members, but AI is also kind of giving me a summary right here, as soon as I pull that up.
It’s telling me all of the activity that’s happened, 11 confirmed members. The opportunity has been in qualification since April. James Smith. CRO, last engaged in November 2025. All of this great information, including advice on next steps, you know, I need to qualify this CRO as an actual member of the journey, but I can also use AI to say.
Hey, you know, who should I reach out next to in this journey? So I’ll just… I’ll just send that command through, and it’s gonna analyze the entire journey the agent is, and all the actions that have happened, the signaling. The adding of the members. Exactly.
And it gives you a clear picture of the journey, so you can take the right next steps with confidence. Exactly. So there you have it. The journey is active, with significant momentum from marketing and sales team. You have a 27, you know, 275K opportunity in qualification, and the buying group is highly engaged. who to reach out next.
And it gives you those, personas and why. And again, following up with some next steps. So, I apologize for the Charlie mishap, but that’s what I wanted to show you today. Thank you very, very much. See if there are any questions.
Hilary Terrell:
Awesome. I’ll pull our slides back up.
Hugh Walton:
I can help walk through that. Let me know.
Hilary Terrell:
Hashtag when Agents Go Rogue. Charlie is a third presenter today. Alright, let me go into full screen. Just a couple things I wanna, reiterate on some of the things that Hugh was showing. First was our scheduling capability.
We were showing literally live on our website, the agent that we use, named Charlie, and allowing that to automatically schedule. You can use LeanData, basically bring your own AI SDR agent. So we were showing qualified, we’ve got customers doing it with Agent Force, OneMind, Expertise, but we have an MCP where you can hook it into any Agentic SDR.
that you are using within your go-to-market motion.
We then looked at the RevOps persona, so that was showing our orchestration product and our flow builder, and then finally showing that journeys and this screenshot that you see here, helping the agent, helping the seller make sense of what was going on with that customer journey by way of, this agent.
So we have time for questions, but Julie, up to you on how you want to use the remainder of the time.

Julia Nimchinski:
Yeah, phenomenal presentation. Thank you, Hilary and Hugh. And we have quite a bit of questions here. So, first one is more, speaking to the philosophy of, you know, like, how you approach agents internally at LeanData. Who should actually own them? And the outcome. So, is it RevOps, IT, Revenue Leadership, or individual business functions?
Hilary Terrell:
that is… like, if I had the correct answer, if there was a correct answer, I would not be on this… on this webinar. I would have solved everything. We actually surveyed our entire customer base to see… just to better understand where it lives for most organizations. The, like… In summary, it lives in a multitude of places.
What we heard from the vast majority is that there is a centralized AI Council, that’s helping to make sure there isn’t duplication of workflows or agents. And obviously RevOps is a key stakeholder in that cross-functional, like, AI, AI Council. Outside of having the Council, the place, or the top two, were RevOps and IT.
Julia Nimchinski:
Thank you. And… let’s see, one more question, what should companies fix before allowing AI agents to take An actual action in production, autonomously.
Hilary Terrell:
Sorry, can you repeat that question?
Julia Nimchinski:
Yeah, so what should companies fix first before allowing agents to act autonomously?
Hilary Terrell:
Yeah, I mean, to oversimplify, I think it goes back to those three pillars that we were talking about. You want the agent to have the correct data. to work from, and it depends on what type of agent it is, right?
Like, we’re, in this context, talking about go-to-market data, so we want to make sure anything that’s flowing through a customer journey, that that data is clean. The second is the process. Like, what do you want that agent to do?
And does the agent have the context of what’s been happening on the marketing side, the sales side, and the post-sales side? I think from a philosophy perspective, the way that we’re building it within LeanData from a product perspective is that it has both the rules-based and the autonomous.
So when we put different information into a graph, or if you’re, you know, recall what Hugh was just showing in the AI assistant, it’s a human prompting it. It’s a human taking the action based on the recommendation versus fully hands off the wheel. So it’s a combination of both the human and the autonomous component.
Julia Nimchinski:
And, Hilary, for those evaluating the category of AI orchestration platforms in the audience. What would be your key differentiator compared to… You can name competitors or not.
Hilary Terrell:
I won’t… I won’t name competitors. I think the biggest places… so, number one, we’ve been around a really long time. LeanData has been around for 12 plus years. We have seen tens of thousands of customer implementations, and so where we shine is complex go-to-market motions that need scalability and governance and trust.
That is what our customers come back to time and time again, and why we have those enterprise logos, like. and Anthropic and OpenAI, who could, in theory, build it themselves, but to maintain it, to make sure it has the right rules and the data, that is our sweet spot, and we’ve been doing it for a long time.
So I’d say being able to manage the complexity of enterprise go-to-market, the scale, and the experience that we have in doing that.
Julia Nimchinski:
Super impressive. Folks are asking, how do you actually avoid, you know, SaaS sprawl being replaced by Agent Sprawl now? Because you have SDR agents, meeting agents, research agents, and, you know, the list goes on.
Hilary Terrell:
No, and now you need an agent for all of your agents. That was a topic at an event that I went to last night. Now people are building the super agent to manage all the agents. I mean, there absolutely is sprawl.
Something else that I heard at this event I went to last night was a new role that people are hiring internally to really be that person who looks across, particularly in a go-to-market perspective, all the different crossover. I think it, you know, depends on the size of the organization in terms of where that shows up.
But again, as we ask our customer base, you have a mix of where agents are coming from. It could be within a software that someone is already using, and an agent has been, you know, added to the stack. You’ve got teams building their own agents.
So, yeah, it is… again, we’re swinging back from… from Wild West to, let’s get our operational guardrails around. So yeah, not… that’s a half-answer to that question, I would say.
Julia Nimchinski:
And Hilary, how do you approach this, you know, just prevailing fear now of, especially in the enterprise? Of giving your own alpha. IP, you know, to name your LLM of choice, but… Yeah.
Hilary Terrell:
In terms of how we are building our own product, or how we’re deploying within our customers?
Julia Nimchinski:
Anything, you know, if you can speak to the harness, governance…
Hilary Terrell:
Yeah, I think the way that we have approached it is to be an agnostic layer. So, there is AI built natively into LeanData. And you can hook it into, if you’re using Gemini or OpenAI, different models, such that, like, you don’t have to be wed to one, and that’s how we think about building our product as well.
Pick the best tool for the job, and be that, like, agnostic, trusted partner. And Hugh, feel free to jump in on any of those.
Hugh Walton:
No, absolutely, I agree with you. You know, it’s bring your own LLM, right? So…
Julia Nimchinski:
Perfect. And, folks are asking about ROI. Like, can you speak to a couple of, you know, I mean, you mentioned amazing examples.
Hilary Terrell:
Yeah.
Julia Nimchinski:
customers, but yeah.
Hilary Terrell:
Yeah, absolutely. In terms… so, a lot of our customers start with LeanData, and a lot of people know us for lead routing and lead to account matching, so more of, like, the acquisition phase of the customer journey.
So, a lot of the ways that customers initially see ROI is in, lead progression, so taking those qualified leads and being able to work more of them more efficiently and qualify them more, and that’s then leading into both increased pipeline and pipeline velocity, so things that enter the pipeline and get closed, and worked more quickly.
a lot of customers don’t realize that we also work post-sales as well, so you can leverage LeanData to optimize maybe a customer onboarding process, to listen for product usage or adoption stats, and be able to intervene before something becomes kind of an account… an account risk. So.
On the post-sales ROI side, we have customers who are seeing, larger and more frequent renewals and expansion opportunities, a reduction in the number of cases and at-risk customers.
So a number of different stats, and I don’t… I should have another monitor, and they’re all on our homepage as well, but those are my… off the top of my head ones I can recall.
Hugh Walton:
to back that up, it’s not just about acquiring leads, right? You… once you acquire them, you have to retain them, you have to service them. So, you know, we use… AI agents to properly assign, you know, customers, prospects to the right people based on whatever market segmentation data that is. Once they become a customer.
what are we doing, after that? Are we creating opportunities after 90 days to make sure CSAT scores are right? Are we creating opportunities in 180 days to. set up cross-sell motions. You know, later on in the year, we might want to start talking about renewal. So, it’s really not just acquiring, but, you know, retaining.
And, and servicing as well.
Julia Nimchinski:
Thank you so much. And last question, what’s on the roadmap? What are you allowed to share?
Hilary Terrell:
Oh my god, it’s like you teed up my last slide. The roadmap is, no surprise, it’s pretty Agentic.
So, we’re unveiling a ton of new capabilities at our event the first week of October, so would love to see folks there, but think about all the ways in which people are using LeanData today, but allowing it to be even more autonomous, because it’s running on top of the rules and the logic that you’ve built. So, that’s what I can hear.
Julia Nimchinski:
Thank you so much. Phenomenal session. And yeah, that’s a wrap for today. We had 10 live demonstrations, 10 very different approaches to autonomous Revenue Agents, and to folks asking, you can access all of the sessions on hardscale.exchange, I believe in 24 hours. And we’ll see you on our AI Summit in September 8th. N?
And the topic will be announced soon. Any closing words, Hugh and Hilary?
Hilary Terrell:
No, thank you so much for having us here.
Hugh Walton:
Appreciate the invite. Thank you.
Julia Nimchinski:
Thank you so much. Decent.