End-to-End Excellence: Transform Your Customer Lifecycle with Breeze Agents

Good morning. Happy to see you. My name is Charlene and as you can hear, I'm French and I'm a professor for HubSpot Academy creating content for our French customer.

Hi everyone, I'm Gigi. I'm here to represent HubSpot Academy as well, but from the German speaking market. And thank you all so much for coming to our session on end to end excellence and how to transform your customer lifecycle with Breeze agents.

Well, first, thank you. Because we know it's 3pm so it's basically nap time still you here. So we will try not to make you sleep too much hopefully.

Let's see, let's see. The coffee stations are downstairs, so let's actually jump straight in. What are we going to have a look at today? So we started off with a little introduction check.

Yes. Then we're going to present you five Breeze agents that will lead us through the whole presentation.

And then next up we have prepared a kind of AI generated customer story. But we'll bring those agents to life just to show you what you can really achieve with them.

Then we will give you a kind of implementation roadmap.

And then finally we'll close off with some resources and next steps. And we'll have a little Q and A at the end as well. So if you have any questions, they will be answered. So you've just heard from us, now we want to hear from you. Please raise your hand if you or your teams do more than two manual tasks a day. Anyone else?

Great, that's a lot. And maybe three or four.

Yeah. Are we still there? Keep them up if it's four. Okay.

You'll see a lot of hands.

Actually if you look around you, you're not the only ones. And that's exactly why we're here today. To help you save up on those manual tasks, save some time. And for that let's introduce your Breeze agents.

So yes, today we will focus on five agents. You can call them your Avengers team, your Justice League, depending if you're more Marvel, more dc. We, we don't care. We're really open minded with gg. Anyway, first we wanted to introduce you to Breez Marketplace. So maybe you've already had a look at it, but this is really your place to go anytime you want to find a new agent. You can find pre built agents or you can customize them depending on your needs.

Okay. So the first agent that we want to introduce to you today is the prospecting agent. Think of it as that BDR that is always first to the office Always last one to leave, never needs a single coffee break. I don't know how, but you know, AI and he always researches customers properly. And it like really digs deep into both contacts and companies. We're talking like CRM history, website visits, job postings, even recent news. So it's like someone on your team who's been following each of your accounts for years, except that it does everything at the same time and deals with hundreds of prospects at the same time. And it takes seconds or minutes. The great thing about it is it operates in over 200 languages. So whether your prospect is based in Stockholm or in Lisbon, it speaks your prospect's language. And not only that, it also speaks your language because you'll be able to set it up with just your value, props, your pain points. It develops an understanding of that. So. So it's like really having an additional team member.

Then we have the customer handoff agent. So this agent is really a game changer when it comes to accounts moving between team members. I'm almost sure you know that problem. Anytime you have an account taken care of by a new team member, then you lose all the context. So everything that was said during the sales process, you, important emails, important deals, information, they are all lost. And what is good with this agent is that it automatically creates a complete handoff report so you can find key contacts and their roles. A summary of deals, information, transactions, next steps. You can also have recent activities. And finally the agent can flag any risks or important items to address. Also, you have no manual work needed. So the new owner, sorry, has to start normal from scratch, but is really in minutes.

Next up, the agent we're going to look at today is the customer health agent. It's like having a crystal ball on your team, but instead of having like vague products predictions, you actually get actionable insights from real data from your CRM system. What it does, it automatically assesses the account health by analyzing calls, emails, tickets, engagement patterns, everything. And it collates it for you into really great insights. So that way you won't surprisingly lose your customers. You'll spot the warning signs weeks before their turn. And when it does detect those risks, it's not just saying, hey Houston, we have a problem, it's actually telling you what is wrong and it'll tell you how to fix it so it'll give you next best actions so that you can save your customers before they turn.

Then we have the customer agent. This one is really your new AI teammate. He can handle 65% of customer conversations automatically. So this means for marketing, for instance, it answers questions and guide visitors to the right content. For sales, it can explain pricing, product details and helps buyers decide faster regarding service. Of course, it resolves common issues and it can escalate to the human teams when needed. The magic is that it works 24, 7. And finally, your customers can get accurate answer at any time.

And then finally, I know it's a lot of information, but we'll break it down a little bit more. Finally, we have the knowledge base agent. This is the agent that really makes your knowledge base knowledgeable. It's like having a technical writer on your team who likes reading support tickets for fun because don't they make the greatest bedtime stories? So what it does, it analyzes content gaps by analyzing questions that keep on coming in. And then it cross references that with the existing knowledge base articles that you already have in your business. So that way you won't have to spend all of that manual labor and time on like, hey, have I written that before? Have I not? It'll tell you and it'll give you those warning signals. Even better, it already drafts the articles for you by looking at the tickets and your customer success or customer support people's responses and using those best answers that your team has already solved. The good thing is it gets smarter with every interaction, like ticket views, article views. All of that feeds right back into the knowledge base article to give even better recommendations. So it's basically like your knowledge base improves itself while you're sleeping. Now, we've talked a lot about different agents. Now I'd like to introduce you to zero to hero a customer success story to actually make everything come to life.

So let us introduce to you Dan. Dan Dillon. He's a happy customer of a company named Velocity. Velocity is an organization renting e bikes to other companies. And the last that we can say is that Vlocity is really nailing the customer journey from the beginning to the end.

Hey there. I'm Dan and I've been a happy Velocity customer for about a year now. All right, just a quick note before we dive in. I'm actually AI generated, so technically I don't exist. But don't worry, you go ahead and enjoy my story while I handle some of these so called real human problems. All right, so let's dive into Velocity and why I'm excited about it. It's been a fantastic experience. From the very first day, it felt like they just got me. Every need I had was anticipated and taken care of. It all runs so smoothly. Honestly, I'm not sure how they pull it off. And, you know, if I could somehow get my hands on their magic formula, I'll definitely use it for our business, too.

Okay, so Dan is happy. Perfect. But why? What was his exact problem? And how did he went from being a stranger to a subscriber at Velocity Again, let's hear from him.

I remember the exact day I realized we had to tackle Employee Wellness head on. I checked out the latest survey results, and a thought just hooked me. Our team is really asking for sustainable commute options. We're losing some great talent to companies that have better benefits. Honestly, the idea of bringing in yet another vendor, it just sounds like a nightmare, doesn't it? And finally, it's like the right vendor just found us.

Okay, let's have a look at how the right vendor just found Dan. But before we do that, let's take a little step back and have a look at how things used to go at Velocity. And maybe the situation sounds familiar to you. So initially, we would have reps drowning in manual research. That would be LinkedIn, stalking, googling, website visits, hours spent just to try and figure out, who do I actually talk to? And then when reps did find someone to reach out to, usually, like HR departments, they weren't really knocking at the right doors because they needed people focusing on culture, like culture officers, facilities managers, all that. Then when they did reach out, had that contact, it was just that generic pitch of like, hey, we lease E bikes to companies. So it's like throwing spaghetti at a wall to see what sticks. So the results, we'd have low credibility with our prospects. And they just sounded like every other vendor in their inbox. Deleted, ignored, and forgotten. The response rates showed that they were brutal, even though reps worked their hardest. So it's a classic case of maximum effort and minimum impact. Does this sound familiar to anyone? Yeah, I see some nodding heads. So this is where our prospecting agent comes into play. So here you'll see the setup of how you just make your prospecting agent your prospecting agent. It's five simple steps. It really won't take long. It does seem like a lot of fields to fill in. But all you really need to do is add your website, and HubSpot will pull the information from your website. Now, what I would recommend doing, I don't know. Have you all attended Kip's session this morning? Yeah. Perfect. Remember when he said AI and human again? We want to make sure to keep that human element. Speak to your reps, get their recommendations of like, hey, what Works for us. What message lines and let that flow in so that you can have your additional team member who speaks exactly your voice. Now let's have a look at what happened with Dan and Velocity. This is the overview of the prospecting agent. Let's say we have our sales rep, Tom. This is what he sees when he logs into his HubSpot account. As you can see, Dan is right there at the very top with TechFlow. And the agent identified him as the person to talk to as the head of people and culture at TechFlow. The great thing is the prospecting agent actually understands what that means. He knows that Dan is responsible for everything that's in relation to like employee well being, culture, all the things. So he knows. Yes, this is the person I want to talk to. Now what do we do with him? The prospecting agent already creates an email draft based on the research it does. As you can see here, in that first line it already mentions TechFlow's commitment to carbon neutrality, which it found from its website. Very specific, concrete data, not just some greenwashing. Next thing we see the London office opening again, pulled in from the research the agent does independently. And then it also noticed some job postings that TechFlow had on their website where they mentioned competitive well being benefits for the employees. So it's highlighting that and it's like connecting the dots and really understanding the context of what this company has to offer. So as you can see, it's not just some generic email. This is the type of email your best rep would create after spending two hours on research. And the prospecting agents does this for hundreds of customers at the same time within seconds. Now this is not where it ends. We also have a couple of very personalized follow up emails. Not just a hey, circling back in, checking in. It's really targeted to our prospect. Now the great thing is Dan is actually interested in the employee benefits angle. So he schedules a discovery call and again, it's all happening right here in this thread. If you remember Nick's presentation from earlier, you can modify your CTA when setting up the prospecting agent so you can set up that meeting link. So it's all happening automated. So this is what happens when the outreach actually feels consultative, not salesy. Prospects want to engage. So let's have a look at what happens next. Dan actually signs up for 100 ebikes initially, but during the discovery call the full opportunity is revealed. Techflow has 500 employees split up over three offices with a wellbeing budget of 300 pounds per employee can Anyone do the math. What's the full opportunity here? £150,000 in revenue potential. Isn't that great? And something that's important to note for later. What gets discussed during the call. TechFlow needs maintenance. They need a dedicated onboarding person and ESG reporting capabilities which Velocity offers. And as everything is integrated in HubSpot, everyone works from the same data. It's all stored so that Dan is now ready to be handed off from his sales rep, Tom to his csm, Lisa.

So, as we mentioned during the introduction, we all know that the handoff is a critical moment during the customer journey. And even though customers are tolerant, they also want to feel unique and they don't have time to waste. They don't want to spend too much time explaining all over again what they already told during the sales process. For them, again, the experience was quite perfect and he will explain it to us.

Hey there, I'm Lisa Dan, CSF at Velocity. I don't mean to tip my own horn here, but Dan seemed really happy with our first call after signing with us. Why don't we just hear it straight from him?

Wow.

Lisa, our new account manager, is absolutely on top of things. She knows everything about what we need. And get this. During our first call, she even remembered our ESG reporting requirements. That's just brilliant.

Okay, so what happened before? Before the customer handoff agent, as we said, customer would have to restate everything they already discussed. This was taking a lot of time. But worse, this means that some key information were lost in the process. So customers were frustrated and they having to repeat themselves a lot. Also, some promises made during the sales process might have been forgotten, which means the experience started not the best way. But now we have the customer handoff agent. And as you can see, it's really easy to set up. Actually you can give it instructions such as goals, roles, method, and you can even tell it which data it should look at to get the most important information. You can also test your agents as you can see on the right side and you will be sure that it works exactly as you need. This requires no credits, so you should really test it before going live. For Lisa, the new CSM at Velocity, she will use this agent to get role specific information. So she will learn that for instance, TechFlow has three offices and London is the most important one for Q1 and is also a pilot. Then she learns that they need ESG reporting for their board. The CFO expects quarterly utilization reports and the agent can also give information such as communication information. For instance, Dan prefers teams when he has quick questions. Also, one important thing. Tom, the sales rep. During the sales process, he promised that they would have a dedicated launch manager. And the program they choose includes maintenance and insurance.

So Dan is now with Lisa with his CSM on his customer journey. But suddenly things start looking not so great. Let's listen to him.

So let me break this down for you. At first we were really happy with Velocity, but I had to dig into the numbers. It turns out we only had about 40e bikes being used regularly. That's where my job kicks in. I need to question the return on investment. Plus, a few employees mentioned they were concerned about the battery range.

That's the moment where I worked my magic and came up with the perfect plan. So let's look at what happened before and what the world of a CSM looked like before. So the team at Velocity initially spend hours and hours and hours looking at spreadsheets. While I'm personally a fan, not everyone is and I respect that they would try to like manually track their customer health. Again, spending hours manually tracking data, logins, tickets, emails. By the time they found the root cause, it was often too late. Then once they found the root cause, they were wondering, hey, what's actually the right approach? Should I call? Should I email? What's the solution? So again, time being lost. At the same time, they missed crucial information. That email that mentioned budget concerns buried in threat number 42, that usage drop that wasn't even visible until the quarterly review. So critical data was being missed. And this is where the customer health agent comes into play. Again, super simple setup. You give it instructions like, again, the method, the goal. You can even tell it, hey, look at this data. If you don't know where to start, speak to your team like usual. Bring that human element in and ask them, hey, what information do you usually need to figure out what's going on with your customers so that you can best support them? So let's have a look at what happened with Dan, TechFlow and Velocity. So this is Lisa's overview in the Customer Health workspace in HubSpot, as you can see, TechFlow is right at the top and at risk. HubSpot identified that they have way, way, way not enough usage. Sorry. English foreign language grammar isn't always grammaring. So the usage is currently at 40% out of 75%. There were several tickets that mentioned range anxiety, meaning, hey, I'm not actually sure your product gets me where I need to get. Dan missed his last check in calls and there's absolutely no engagement with the training resources. So what does Lisa do? She uses the customer health agent. She runs TechFlow through it. And this is the information that she gets. And this is a really, really long report. We couldn't all fit it on one page, but these are kind of the most important points. As you can see right at the top, the customer health agent tells her, hey, this customer is at risk. He has a really high priority. It also identifies the root cause confidence issues among staff. So the product is fine. People just really don't know. They're really uncertain about using it. And then you can see here the recommended immediate actions. In this case, it's suggesting to Lisa to launch a peer ambassador program to really bring to life and highlight that everything is fine and that employees really don't need to worry about. And it gets that information based on what worked previously with other customers. So it's not just a guessing game, but it's using the data you have in HubSpot because everything is integrated. So yeah. And what it also does, it proposes an email that Lisa can send to Dan already mentioning all the data. Hey, from experience with similar rollouts, we've seen that a peer ambassador program can help employees feel more supported directly addressing the concerns, not just sending an email with like, hey, how's it going? What's wrong? We know what's wrong and we're contacting Dan with a solution. It also proves the data. One of our recent clients saw adoption rates rise from 42% to 74%. Again, getting very specific, very personalized, and again, already proposing next steps. Scheduling a brief call and giving a time frame for the next 24 hours. So what happens as a result of this email? We have a perfect response from Dan. We get on a call with him, he's actually super happy. After implementing the peer ambassador program, employee confidence rises. The number of tickets mentioning, hey, how far does my bike go? Actually lowers that range. Anxiety is practically gone. And this is how a rescue situation, an at risk customer, goes into expansion. Can I have a quick show of hands? Anyone here in the room who's lost a customer that they could have saved? In hindsight, yeah, exactly. So we all have that similar issue. We're all in the same boat. And hopefully this agent will help you solve exactly that.

Okay, so so far so good, right? I mean, what could happen? What could prevent Dan and all his colleagues from being happy forever with Velocity? Do you have any idea? Because I have one. Support. I mean, what is support if it's not helpful? You can have the best marketing you want. You can have the best sales rep. Amazing customer success manager. What if your support isn't efficient? Well, again, and I'm really sad to say it, I mean, Velocity did really well. And I know it kind of gets boring now, right? They are so perfect. But let's hear what happened.

All right, here's the story.

This bot, I'd say it saved my life.

I was having this E bike nightmare. It just wouldn't start. And get this, I was almost late trying to a super important client meeting. But thanks to the amazing support I received, I was able to restart the bike in two minutes. Just bam, up and running. I was thrilled. The clients were thrilled. It was a good day for everyone.

Great. Thank you, Velocity. What happened before? The problem is, anytime customers add urgent issues, they had any possibility to have a quick answer. Why? Because every issue went to the same queue, whether it was critical or simple. And the teams, they had to answer the same repetitive question all over again and couldn't focus on the real issues, the complex ones. So of course it led to frustration, frustrations for the customers, frustrations for the teams, and exhaustions also. And finally, we had a support that couldn't help anyone effectively. But now we have the customer agent. The customer agent is really easy to set up and you can really give a lot of details and put details into it. First, you choose its personality. So whether you want it to be friendly or casual, it's possible. If you want it to be professional, empathetic, it's also possible. And you can even match it to your brand voice. So every conversation feels really authentic. Next, you feed it with your knowledge. Whether it is your knowledge base, your website pages, your blog posts, anything you have in HubSpot or outside of HubSpot. It can be external URLs, you can upload files, you can really feed with everything you have. Then you have to assign it to channels. And it can be websites, chats, emails, Facebook, WhatsApp. It's up to you. The best part is your agent really knows its limits. If it can answer, it will answer. And it will use your knowledge base to give accurate answer that your customers can really, really trust. If it's not sure, it will ask clarifying questions. And if it is sure that it's not able to answer, then it will scale it to human teams. Finally, the result is your customers get instant accurate answers and you can scale your support for Nikita. Here, what it looks like. So the diagnosis was battery safety cutout after long idle. The solution, step by step reset guide via chat. The result, E Bike working in three minutes and as a backup also sharing nearest E bike location just in case. In the meanwhile, the knowledge base agent notices that support received 95 bike won't start tickets this quarter. And there is a pattern. Monday morning battery issues. Yes, let's have a pause here. Monday morning battery issues. Really? Yeah. Anyway, then it takes actions. Creates new help center article automatically. The result is crystal clear. More than 50, 75% fewer dead battery tickets. The customer agent handles remaining 25% and the support team can focus on real complex issues.

And this is where our story comes to an end. We've come from being a cold prospect to expansion to becoming an advocate powered by agents who help both Tom and Lisa be brilliant at their jobs. This is what end to end excellence looks like.

All right, folks, let's talk about Velocity and what a game changer they've been for our benefits program. Can you believe it? We've got 80% of our team riding E bikes now. And here's the kicker. We slashed our carbon footprint footprint by a whopping 30%. But that's not all there. Employee satisfaction is through the roof at an all time high. Velocity isn't just a vendor to us. They're a real partner and I will keep recommending them as much as I can.

Now, this might have just been a partially AI generated story, but Dan's journey isn't fiction. It's happening right now to companies you using Breeze agents. The only question is, will your customers be this happy tomorrow or will your competition get there first? But don't worry, we've got you covered. We have a little implementation roadmap that's really easy to follow. Charlene, want to kick us off? Yes.

So we have questions for you. If you feel like you're drilling into support volume, then our recommendation is to start with the customer agent. If manual research takes hours per prospect, then you should try the prospecting agent if it's not done already. Do you have a high churn rate if the answer is yes, I think you already know the answer. The customer health agent is your agent. Go. And finally, if you feel like transitions between the teams are a bit messy, then give it a try. To the customer handoff agent.

Thank you. Now, as Charlene and I are here representing HubSpot Academy, we have a little bit of homework for you. So during the next week, I'd like you to pick one agent that we discussed today or another one that you find in the Braze marketplace and find the highest impact use case that Charlene just introduced to really Use that as the basis and focus on that one issue you're trying to solve. Then choose your agent based on that. Document your baseline metrics. So where are we today? And then also keep track of early results. So what kind of things are improving for us? What kind of things are not working? What do we test? So that way you can really get the full benefit of using this agent.

Also, some of the agents we introduced today are still in beta. But here is the real impact of Sunbreeze agents and it's real on customers already. So few numbers. The reps using prospecting agent, they decrease their time spent researching accounts by up to 95%. Customers using prospecting agents engage up to 3.5 times more leads. The customer agent, as we already mentioned during the presentation, solves over 65% of conversations automatically. And finally, the teams using customer agent Resolve tickets 39% faster.

So this is what your business could look like once you start implementing agents. Now we're almost at the end. We're going to go into Q and A very quickly or very soon. In the meantime, I'd like to invite you all to take out your phones and scan this QR code. You can see here, you'll have some additional resources, tutorials, guides, Breeze Marketplace, also our LinkedIn profiles if you want to connect with us. But yeah, we'll give you a minute to do that. In the meantime, if our solutions engineers want to join us, Amin and Carlos, they're in the room with us for some more technical questions and I think we're ready for the Q and A. Yeah, we're ready to hear from you. We have someone just over there. Carlos, Hi.

Thanks for the presentation.

On a health agent, is it possible.

To aggregate up all the feedback that you get from customers so that you get a view of the pain points that they're expressing?

So the customer health agent right now works per customer, but we have other tools in HubSpot that let you do that. So we have the feedback tool that will let you aggregate that. Actually, there is some new feature features where you can use Breeze to aggregate that and I think it's currently still in beta, but there's a feature where you can even look at different themes. I think the tool is called Themes, but please correct me if I'm wrong so that you can look at specific things that you're interested in for your team. Does that answer your question?

Thank you.

Perfect. Thank you. We have another question just here in the front. Oh, thank you. Amin.

Hi, thanks. Customer health agent question again. Is it possible to pull in third party data for the customer health agent to make a more holistic assessment of your customer's health. Because most of our most pertinent data wouldn't be stored within the CRM, it would be within our back office systems or even within like Excel spreadsheets. So I don't think you'd be able to get like an accurate picture of the customer health just from looking at the CRM interactions.

So at the moment, as far as I know, it works with the data you have in HubSpot and available in HubSpot. I don't know if that's going to extend and like what the integration pipeline looks like. For now, it's mainly the data that is stored within HubSpot. I don't know if you have like any workflows or so that you can implement to kind of get that data into HubSpot somehow. For now, it's HubSpot.

Thanks.

Thank you. Yes, we have someone there in the back. Thank you. Also, can we give a quick round of applause for Amin and Carlos for doing all of this. Thank you so much.

Yeah, thank you very much.

I had a question.

These AI agents are very specific. Specific for certain use cases. Will there be the possibility to create.

Your own agents in the future?

Yeah, I think so. Yeah, please, Because I think you can already do it.

Is it working?

Yeah, yeah, yeah.

So yeah, they are like the.

Actually the customer handoff agent is based off of the framework of agents that.

You can build yourself. So, yeah. Thank you. All right, any other questions? Yes, we have someone here right in the middle. Thank you. When you run these agents, where do the reports end up? Do they automatically associate to the company that it references or do they live somewhere else? So when you're in Breeze Studio, you will find all of your agents, or depending sometimes you'll find them in the navigation. So you'll directly go into those agents and that's where you'll find it. So for example, for the customer health agent, you would go into the agent that lives in Brief Studio. That's where you'll find it. And then you just run it through and then it'll be right there. So it won't be stored on the company or contact records, but it'll be directly in the agents in most cases. Does that answer your question? I think so. So it sounds like when the agent runs, the report that it generates won't automatically attach itself to the associated company. No, but you can download most of the reports, share them with your teams and you can set up those Associations. I'm not sure if you might be able to implement workflows, but for now, they live within the agents. Thank you. Thank you. Any other questions? Oh, we have someone right here.

Hey, what's the accuracy rates of your AI agents? And can we do anything to improvise that accuracy? Can we train those AI agents further, see the evaluations where it's probably not more deterministic, and improvise it further?

So I can't give you specific numbers on the accuracy. What I can tell you is you can customize those agents so that they meet your business needs. So when we looked at all of the different setup that we saw at the beginning, you have room to test all of these agents to make really sure they give you the accurate responses. Now, they don't tend to make up answers. I know AI does that sometimes, but you will give it the resources that it needs to train them. So, for example, if we look at Customer Agent, you upload all of your knowledge base. You can include your website URL, you can upload documents, that's the resource that it uses and it quotes those resources. Even when we look at the prospecting agent, the research that it does, you can look into the kind of things that it. Or you can look into the data just to verify, hey, is this actually accurate? So that way you can build that trust in those agents? Does that answer your question? Yep, to an extent.

Thank you.

Perfect. Thank you. Any other questions? Yes, right in the front.

You may have already mentioned, but earlier in the spotlight there was mention of Gemini, OpenAI and Anthropic as being three that you could use. I haven't seen the reference.

Is there any documentation available as to.

How you'd configure that?

Is that applicable to the agentic frameworks.

We'Re talking about here or.

Yeah.

Yes. So not sure to understand correctly the question. Are you talking about connections between those systems or.

So obviously you've got agents here that we're discussing in terms of the different capacities they have and what they do.

So Customer Health as an example, can.

We choose to use Anthropic's CLAUDE model with that agent, or is that agent specifically set up with OpenAI?

Oh, okay. So correct me if I'm wrong, but you can find all the models used, because agents, they use different models already from anthropic, from OpenAI and it's really specific. The AI in HubSpot is using different models depending which tools you're using. So you have a website you can have a look at and please, you can Add any information if you want.

I mean, once I. Oh, yeah, sorry, yeah.

So just to add to that. So there are two ways where you can integrate with an external LLM. So if you're using your own ChatGPT or OpenAI, the first one is to use HubSpot MCP server that you connect to an McP client like ChatGPT or Claude or Gemini, and basically you're going to be interrogating HubSpot data with natural language or doing deep research on that. So that's one way that you can kind of integrate with your own LLM. The second part is in the workflow. You're going to see an action called Use Custom LLM. So if you have, for example, a specific project on ChatGPT that's consuming data that lives outside of HubSpot in the workflow, you can send the prompt to that LLM, get the answer back, and then use it in the rest of the workflow. So these are kind of two things that already allow us to interface with 3rd LLMs, but there are other things on the roadmap. And as we continue through that roadmap, does that answer the question?

Yes.

Thank you, Hasi Babu.

Amazing. Thank you so much. We might have time for one final question before we say thank you. And I've just clicked that too early. One last question. All right, going once, going twice. In that case, thank you so much, everyone for your time.