HubSpot Breeze Agents Explained: Use Cases, Setup, and Real Impact

Breeze Agents are flipping the script for marketing, sales, and support teams by delivering AI automation that actually gets stuff done. These are not just chatbots or trivial assistants. They act like digital team members embedded in HubSpot, handling the busywork in content creation, social media, prospect outreach, and customer support. If you’ve ever thought, “If only I had an extra pair of hands,” Breeze Agents might be the answer. They free your real team to focus on strategy and creative work while the “AI teammates” tackle the repetitive tasks. The result? More leads, faster sales cycles, and happier customers. All without adding headcount.

In this guide, I’ll demystify HubSpot’s Breeze AI Agents, what each one does, smart ways to use them, how to set them up, and the real-world impact they’re making for SaaS companies. By the end, you’ll see how these tools can empower a lean RevOps team to achieve big results. Let’s dive in.

What Are HubSpot Breeze Agents (and Why Should SaaS Marketers Care)?

If you’re hearing about HubSpot Breeze Agents for the first time, think of them as AI-powered experts living inside your HubSpot platform. Each agent specializes in a different area: content marketing, social media, sales prospecting, customer service, or knowledge base management. Importantly, they don’t just chat or make suggestions; they can take action and automate multi-step tasks that would otherwise eat up your team’s time.

In plain terms, Breeze Agents are like digital coworkers. They’re deeply integrated with your HubSpot CRM and content, so they work with full context. In fact, as HubSpot’s Head of AI, Nicholas Holland, put it, “they have all the context from the entire customer journey… they’re built to work with humans, to extend the power of teams. That means they use the data you already have. From contacts and website activity to emails and support tickets to make informed decisions. They operate within your existing Hubs (Marketing, Sales, Service, and Content) rather than in a silo.

Why should SaaS marketers or HubSpot users care? Because this translates to real leverage. Imagine finally having help with those tedious tasks that never make it to the top of your to-do list. No more staring at a blank content calendar or struggling to personalize hundreds of outreach emails. A Breeze Agent can draft a blog post for you, queue up a week’s worth of social posts, follow up with leads, answer customer FAQs 24/7, and even write new help articles when it notices something’s missing. All this happens in the background, quietly saving hours. For a Marketing Ops or RevOps manager, Breeze Agents mean scaling output and improving customer experience without needing to hire more people or outsource work. In short, you get to punch above your weight, achieving big-company productivity with a small-team budget.

In the sections below, I introduce each of the five Breeze Agents, Content, Social, Prospecting, Customer, and Knowledge Base, through common SaaS marketing and operations challenges. For each agent, you’ll learn what it does, how it helps (use cases and benefits), and how to set it up. Let’s explore how each “AI agent” can become your new digital teammate.

Breeze Content Agent: Tackling Content Bottlenecks with AI

Every SaaS marketer I know has hit a content bottleneck at some point. Blog posts, case studies, product pages, e-books. The demand never ends, and small teams can only produce so fast. I remember a quarter where we needed four new case studies, a dozen blog posts, and a major landing page refresh all at once. We were drowning. The Breeze Content Agent felt like a lifeline for those moments.

What the Content Agent Does

Think of the Content Agent as your on-demand AI content creator living in HubSpot. It can generate complete, formatted content pieces in your brand voice, using both your prompts and your existing CRM/data for context. It handles a range of content types, including:

• Blog posts: It can brainstorm topics, draft titles, create outlines, and even write full blog drafts tailored to your audience (complete with suggestions for images and SEO considerations).
• Landing pages: It produces landing page copy and can pull in your existing HubSpot forms or meeting links, giving you a ready-to-go page that just needs your polishing.
• Case studies: Provide a few details or data points, and it will draft a compelling case study to showcase your product’s value.
• Podcast scripts: Surprisingly, it even helps generate podcast episode content. From outlines and talking points to show notes. (Content marketing is no longer just writing!)

All of this is done in alignment with your brand voice and style guidelines, which you can “train” the AI on by feeding it examples. The first time I tried Content Agent, I had it generate a ~1200-word blog post about a new feature release. It suggested a title, gave me an outline, and in about 5 minutes I had a full draft that would’ve taken me days to write from scratch. Was it perfect? Not exactly! I spent another hour editing to add our unique flair and ensure accuracy. But an hour of editing versus a full day of writing is a huge win. The Content Agent essentially kicks off the first draft for you, so you’re never starting from a blank page.

How It Helps SaaS Teams

For lean marketing teams, the Content Agent slays the content backlog dragon. Instead of being a bottleneck, you become a content engine. This means more frequent blog posts (which boosts SEO and inbound leads), up-to-date case studies for sales to use, and landing pages ready to launch new campaigns quickly. Consistency and speed in content creation ultimately drive growth, more website traffic and leads from your content marketing, plus more sales enablement materials to fuel your funnel.

Because the AI uses your CRM data and learns from your past content, the output is surprisingly relevant. It’s like handing a junior writer all your best assets and saying, “draft this, quickly.” In fact, HubSpot says Content Agent helps marketers create high-quality, relevant content at scale based on business context. From my experience, that rings true. The bottom line: your team can accomplish a lot more content creation without burning out, which is a competitive advantage for any SaaS marketing org.

(A quick note: Always review the AI’s draft. The Content Agent is like a tireless intern. It can churn out a decent first draft of anything, but you’re the editor-in-chief who fine-tunes it. The more guidance you give it upfront, the less you’ll need to edit later.)

Setting Up the Content Agent

Getting started is refreshingly simple (no coding or complex configuration needed). If you have HubSpot Content Hub Professional or Enterprise, you already have access to the Content Agent. Here’s how to launch it:

1. Navigate to Breeze Agents: In your HubSpot portal, go to the Breeze section (you might find it under your Content tools or in the main navigation if updated). Select Content Agent and click “Launch.”
2. Choose a content type: You’ll be prompted to pick what you want to create, e.g. Blog Post, Landing Page, Case Study, etc.
3. Provide input/prompts: The agent will ask for a few details to get started. For a blog post, it might ask for the topic or target keywords; for a landing page, your goal or audience; for a case study, some key points about the success story. The clearer and more specific you are, the better the draft it will produce. (For example, give it your buyer persona info or key messages if you have them.)
4. Generate and refine: Hit the generate button and let the AI do its work. In seconds, you’ll get a draft. Review it directly in HubSpot’s editor. You can accept it as-is, or more likely, spend some time editing and injecting your brand personality. The agent learns from these tweaks over time.

HubSpot’s knowledge base has detailed walkthroughs (e.g. how to generate a blog post with AI) if you need more guidance. But generally, it’s straightforward. Pro tip: provide as much context as possible when prompted, e.g. tone (“professional but witty”), target persona, any must-include points. This reduces editing later. And remember, you maintain full control: you decide what actually gets published.

Breeze Social Media Agent: Say Goodbye to Social Media Neglect

Raise your hand if managing your company’s social media often gets pushed to the back burner 🙋‍♂️. In many B2B SaaS teams, social media is important for brand awareness and engagement, but it’s hard to maintain a steady drumbeat of content. I’ve seen Twitter (X) accounts of great startups go silent for weeks, or LinkedIn pages that only post when there’s a major announcement. It’s usually not due to lack of caring. There’s just too much else to do. The Breeze Social Media Agent is designed to solve this by acting as a tireless social media manager that never runs out of ideas.

What the Social Media Agent Does

The Social Media Agent auto-generates social posts across your connected accounts (LinkedIn, Facebook, Instagram, X/Twitter, etc.) in your brand voice. It doesn’t randomly create content. It’s guided by your marketing goals and real data. Specifically, it can:

• Draft platform-tailored posts: It creates suggested posts for each platform, adjusting style and length (short, snappy text for Twitter; a more professional tone or longer form for LinkedIn, etc.) while keeping messaging consistent.
• Recommend posting times/frequency: Based on when your audience is most active (it learns this from your engagement data), it suggests an optimal schedule so you post consistently at the best times without needing a social media “guru” to figure it out.
• Use your engagement data to improve: It knows which of your past posts got high engagement (likes, comments, shares) and which fell flat. Using that, it can double down on topics or formats that work and avoid the duds. This means over time, the content ideas it proposes align with what resonates with your audience.
• Generate accompanying images: Need a visual to go with the post? The agent can even create simple graphics or imagery using AI, thanks to HubSpot’s integrated image generation (“Regenerate image” feature). This is a huge plus if you don’t have a designer handy for every social post.

In practice, I used the Social Agent to keep a SaaS client’s social feeds active during a product launch when the team was too swamped to compose daily posts. We set the agent’s parameters (essentially told it the campaign theme, target audience, and the key hashtags/topics), and it started queueing up post suggestions. Each morning I’d spend 10 minutes reviewing its suggestions, tweaking a word or two, and then approving/scheduling them. It was like having an assistant prepare the drafts so I just had to finalize. It felt great to know our social media was “humming along” in the background, driving engagement, while we focused on the launch event and other pressing tasks. No more guilt about neglecting followers or missing prime posting windows: the agent had that covered.

How It Helps SaaS Teams

The obvious benefit is time saved and consistent presence. Instead of scrambling for something to post (or, worse, letting your accounts go dark), you have a steady stream of content going out. Consistency on social media keeps your brand visible and builds trust over time. It can even generate leads or signups when you share the right content (e.g. blog posts, case studies, event promos) at the right time. Essentially, the Social Agent ensures you’re always on socially, without the stress.

Another big benefit is that it makes your social strategy more data-driven. Because the agent bases its suggestions on engagement data (“what works”), you’re likely to see improved metrics – more likes, shares, comments – as it optimizes content and timing continuously. That level of analytic adjustment would be hard for a human to do manually at the same pace. For SaaS companies, a lively and well-targeted social presence can help with everything from attracting talent (job seekers check socials!) to nurturing prospects who follow your content. It keeps you in the conversation.

A pleasant surprise I found: the Social Agent can maintain brand voice quite well. I was initially worried an AI might create off-brand messaging, but by feeding it examples of our past posts and a brief on our tone (professional but friendly), the posts it generated were on point. Of course, I still review everything. Think of the AI as your social copywriter, not your approver. HubSpot smartly allows you to approve or edit each suggested post before it goes live, so you remain in control of your brand narrative.

Setting Up the Social Agent

The Social Media Agent is available for Marketing Hub Professional and Enterprise users. To enable it, head to your HubSpot Marketing tools (or the Breeze > Agents dashboard if available). Here’s how to get started:

1. Launch the Social Agent: In HubSpot, go to your Social media management section. When creating a new post, look for an option to use AI suggestions (this is effectively the Breeze Social Agent at work). If you have the Breeze Agents dashboard, you can launch it from there as well. (If it’s newly released, you might see a “beta” tag or need to ensure your account has the feature enabled.)
2. Connect your social accounts: Make sure all the relevant social accounts (LinkedIn, Twitter, Facebook, etc.) are connected to HubSpot if not already. The agent will post to any account you’ve integrated.
3. Define posting parameters: You’ll likely be prompted to specify some settings like your target audience or persona, themes or topics to focus on, and how frequently to post on each channel. For example, you might tell it to post 3 times a week on LinkedIn and 5 times on Twitter, focusing on topics like product tips, industry news, and case study highlights.
4. Review and approve content: Once active, the agent will begin generating post suggestions and a calendar. Plan to spend a few minutes (daily or weekly) reviewing these suggestions in HubSpot’s social publishing tool. You can edit text, change imagery, or delete suggestions that don’t fit. Then approve and schedule the posts. Over time, as you approve/tweak content, the agent learns your preferences.

Because this feature was in beta for a while, you might need to check HubSpot’s latest guidance if you don’t see it immediately. But as of now, it should be generally available. Once it’s up and running, you’ll wonder how you managed social without it. I certainly do. It turned what used to be a stressful, ad-hoc chore into a smooth, automated workflow.

Breeze Prospecting Agent: Build a Qualified Pipeline (While You Focus on Closing Deals)

Let’s shift to the sales side of things which, as a marketer, might still be part of your world in a SaaS company (Marketing and Sales alignment is core to RevOps after all). Pipeline development is often a team sport between marketing and sales. Marketing generates the MQLs, but sales (SDRs or AEs) have to research those leads, personalize outreach, and keep the funnel moving. In smaller SaaS outfits, I’ve seen marketers themselves do a lot of the prospecting or nurturing, or founders jumping in to email prospects. It’s time-consuming work that demands a personal touch. The Breeze Prospecting Agent is like an AI Sales Development Rep that loves doing the tedious research and initial outreach for you.

What the Prospecting Agent Does

This agent automates the early-stage sales outreach process by researching contacts/companies in your CRM and then executing personalized email sequences to engage them. Think of it as a smart hybrid of a researcher and a salesperson that runs inside HubSpot. Here’s how it works and what it can do:

• Research your leads: The Prospecting Agent combs through your HubSpot CRM data (like recent form submissions, website page views, past email or call logs) and even looks at external signals if available (intent data, LinkedIn info via integrations) to gather context on each prospect. It basically prepares a mini dossier on what each lead is interested in or how they’ve interacted with your company.
• Write personalized emails: Using that research, it drafts outreach emails tailored to each prospect, far beyond a generic template. For example, if Prospect A visited your pricing page and downloaded a whitepaper on “Data Security”, the agent might draft an email highlighting how your product helps with data security, referencing that pricing page visit (“I noticed you were looking at our pricing for the Teams plan and thought you might have questions about our security features...”). These emails aim to feel one-to-one and relevant, increasing the chance of a response.
• Sequence and send follow-ups: You can have the agent enroll contacts into an automated sequence (cadence) that you define. It will send the first email, wait a few days, send a follow-up (maybe with a different angle or a case study attached), and so on. All automatically. You choose the rules: for instance, stop if the person replies or if they click a link. You can opt to manually approve each email before it goes out, or let the agent run autonomously for scale. Even in autonomous mode, you set the boundaries and content guidelines, so it’s not going rogue.
• Target account research: Beyond individual leads, the Prospecting Agent can research target accounts as a whole and identify key contacts to reach out to. This is great for account-based marketing/sales. Instead of a sales rep spending an hour on a company’s website and LinkedIn to map out stakeholders, the agent can summarize that info and even draft outreach emails to multiple contacts at that company.

In practice, I used the Prospecting Agent for a client who had a large influx of trial sign-ups that weren’t being followed up thoroughly. Our small sales team was overwhelmed, and frankly a bit burnt out writing “Just checking in” emails every day. We set up the agent to automatically enroll new trial sign-ups into a multi-touch sequence: Day 1 -> welcome email, Day 3 -> check-in with a relevant case study, Day 5 -> “Let’s chat” invite, etc. All emails were personalized based on what the user did during the trial. The results after a month were impressive: a significant uptick in email engagement (opens and replies) because the emails were more on-target. One of our sales reps even joked that the AI “writes better emails than I do on a Friday afternoon” - and he was kind of right. Importantly, this freed the reps to spend time on the hottest leads (those who replied or showed strong intent), while the agent gently nurtured the rest. Our pipeline of Sales Qualified Leads (SQLs) grew about 30% that quarter, which I partly attribute to the agent ensuring no lead fell through the cracks.

How It Helps SaaS Teams

The Prospecting Agent directly contributes to pipeline growth and efficiency. By automating the grunt work of prospect research and initial outreach, your team can reach more prospects in a personalized way without working longer hours. More leads get touched and nurtured than would otherwise be humanly possible. In B2B sales, timely and relevant follow-up is everything. A prospect is far more likely to convert if they feel you truly understand their needs and are prompt in addressing them. This agent helps deliver that personal touch at scale.

It’s especially powerful for lean teams or times when your sales staff is at capacity (or out on vacation -> the AI doesn’t take vacations!). For marketers, it ensures that the leads you worked hard to generate are actually followed up properly, increasing the ROI of your campaigns. There’s also a data advantage: the Prospecting Agent uses HubSpot’s Smart CRM data and any integrated intent data to prioritize who to contact and when. For instance, it might prioritize a lead who’s visited your pricing page 3 times this week (a strong intent signal) over one who just downloaded an eBook. That kind of dynamic prioritization is hard to do manually at scale. HubSpot says the Prospecting Agent “helps build a more qualified sales pipeline,” and I’ve seen that firsthand. Quality improves because outreach is targeted and timely, and the agent can even qualify leads by asking questions or gauging responses. Essentially, you’re scaling those top-of-funnel conversations without blasting generic spam to everyone.

One more benefit: the consistency and persistence of follow-up. The AI won’t forget to send that second or third email, whereas a busy rep might. It ensures every lead is nurtured to the extent possible. This means fewer leads slip away quietly. When a prospect does show interest (opens, clicks, replies), the agent can alert a human rep to step in, making sure hot opportunities get a personal touch at the right moment.

Setting Up the Prospecting Agent

The Prospecting Agent comes with Sales Hub Professional or Enterprise (make sure you have one of those tiers enabled). To set it up, follow these steps:

1. Enable the Agent: In your HubSpot account, navigate to Breeze > Agents or the Sales Hub settings where Prospecting tools are. If it’s your first time, you may need to opt-in or turn on the beta. Once enabled, select Prospecting Agent and click to create a new agent or sequence.
2. Define enrollment criteria: Decide which contacts should be fed to the agent. For example, you might say “Enroll any new lead who fills out Contact Us form and has Company Size > 50” or “Enroll all trial sign-ups from this month.” You can also start with a static list of contacts or even manually enroll a few contacts as a test. The key is to target who you want the AI to reach out to (you don’t want it emailing everyone arbitrarily).
3. Set up the outreach content: Next, configure the sequence or rules. The agent might have default email templates it suggests, but you should guide it. Input the goal of the outreach (e.g. book a demo, educate about feature X) and any specific value props or content to include. The agent will generate personalized email text for each step. Review the first few emails it proposes to ensure the tone and messaging are right. This is like training your AI sales rep, spend time upfront so it represents your company well.
4. Activate and monitor: Turn the sequence on and let it run. HubSpot provides a dashboard for the Prospecting Agent where you can see who’s been emailed, who opened or clicked, and who replied. This is similar to a marketing email or sequences tool dashboard. Monitor performance here. Importantly, when a prospect replies or takes a desired action, make sure a real sales rep follows up immediately (HubSpot can automatically assign the conversation to an owner). The agent will typically pause or stop for that contact once human interaction starts.

HubSpot’s knowledge base article on using the Prospecting Agent (which was in beta) can provide detailed guidance if you need it. My advice: start with a small batch of leads and “babysit” the agent initially. Once you trust its output and logic, you can scale up to larger lists or even set it to auto-enroll new leads continuously. One of my clients calls this their “AI sales intern”. It handles the tedious tasks, but the sales team still oversees and mentors it, so to speak. That’s the right mindset: it’s there to assist, not fully replace your sales development strategy.

Breeze Customer Agent: Scale Customer Support Without Sacrificing Quality

Customer support is where a lot of SaaS companies either shine or struggle. Quick, helpful support can turn users into raving fans, while slow or unhelpful support can drive churn. The challenge is scaling support when your user base is growing but your support team isn’t. I’ve consulted for companies where one support rep was handling hundreds of tickets a week, which is an impossible load leading to backlogs and burnout. Enter the Breeze Customer Agent, which I affectionately call the 24/7 support AI. It’s like adding a virtual support rep who works all hours, instantly answers common questions, and never loses patience.

What the Customer Agent Does‍

The Customer Agent uses AI to answer customer questions in real-time via chat or messaging, using your existing knowledge base, website content, and any documentation you feed it. It integrates with HubSpot’s Chatflows (the live chat/chatbot feature on your site) and can also work in other channels like Facebook Messenger or WhatsApp (as of the latest updates). Here’s how it functions:

• Instant Q&A from your content: Once trained, the agent can field a user’s question in the chat widget (e.g. “How do I reset my password?”) and provide an immediate answer pulled from your knowledge base article or a snippet from your documentation. It responds in a conversational tone, so it feels like chatting with a helpful human, not just getting a link to a FAQ.
• Learns from unstructured data: Beyond your formal knowledge base articles, it can ingest other resources like past support tickets, email threads, or PDF manuals (up to 1,000 pages of content). This means even if not every question is documented neatly, the agent can learn from how your support team has answered things in the past. I saw a client benefit hugely from this: they hadn’t written down every Q&A, but by letting the AI read through a trove of old support emails, it picked up a lot of “tribal knowledge” and was able to use that in answers.
• Confidence-based routing: The Customer Agent is smart about knowing its limits. If it’s confident it knows the answer (based on the data it has), it will answer the user and often cite the source (“According to our FAQ page, you can reset your password by…”). If it’s not confident or the question is too unique, it will seamlessly escalate to a human agent or create a ticket for your team. In other words, customers get help fast for common stuff, and your team only gets involved when the AI isn’t sure or when a personal touch is needed. This avoids the pitfall of chatbots giving wrong answers. It won’t guess; it hands off when appropriate.
• Handles simple actions: In recent versions, the Customer Agent can perform simple tasks if connected to certain systems. For example, it might trigger a password reset email or check an order status if your systems allow that. It’s not going to do complex account changes (and you wouldn’t want it to without oversight), but those little actions (like “resend confirmation email”) can save a ton of time for both customers and support reps.

How It Helps (Outcomes)

The impact of the Customer Agent can be dramatic. For routine questions (“How do I add a user to my account?”, “What’s your uptime SLA?”, “Where do I find the API docs?” etc.), customers get instant answers instead of waiting hours (or days) for an email reply. This boosts customer satisfaction and can even improve retention, because users feel taken care of and can solve issues quickly on their own. Internally, your support team suddenly has a lot more breathing room. They can focus on the thornier, complex issues or spend time proactively helping customers, rather than drowning in repetitive tickets.

HubSpot shared that companies using the Customer Agent have resolved over 50% of support tickets via AI and spent ~40% less time to close tickets on average​ One real-world example: Youth on Course (a nonprofit) saw a 75% increase in support ticket volume but actually improved response time by 16% and raised customer satisfaction by 7% after implementing the Customer Agent (these numbers were shared in HubSpot’s case studies). In other words, even as they grew and got more inquiries, their support got faster and better. A feat only possible by offloading the common questions to the AI agent.

From my personal experience, a SaaS client of mine added the Customer Agent to their app’s live chat. Within the first month, the agent was handling about 60% of all chat queries without human intervention. That’s hundreds of answers that the team didn’t have to type out. We kept an eye on the conversations (you can review what the AI is telling people via HubSpot’s conversation logs) and were pleasantly surprised by how accurate and helpful the answers were. And when the agent wasn’t sure, it would say “Let me connect you with a team member for that” and ping our support team. Exactly what we’d want it to do. The support reps started jokingly calling the AI agent “Brad” (don’t ask me why) and would say “Let Brad take first stab at it.” Before long, “Brad” the bot became an integral part of the team!

Setting Up the Customer Agent

This agent requires Service Hub Professional or Enterprise (since it ties into service features like chat and knowledge base). To create one, go to Service > Chatflows in HubSpot, and you’ll likely find an option to add a Breeze Customer Agent to a chatflow. HubSpot’s setup guide (see “Create a customer agent” in their knowledge base) provides full steps. In brief, the process is:

1. Choose content sources: Select which knowledge base articles, web pages, and other documents the AI should learn from. Make sure these sources are up-to-date and accurate. If your knowledge base is thin, you can supplement by uploading a document of common Q&As or letting it digest past support emails. The more quality info you feed it, the better it performs.
2. Configure the chatflow: Decide where this agent will appear, e.g. on your website’s help chat widget, or inside your app. Set when it should answer and when it should pass to a human. You can customize the hand-off messages (like “I’m going to connect you with a specialist now…”). Essentially, define the rules of engagement: does the AI greet first? Does it handle everything on the help center page but not on sales pages? Tailor it to your support strategy.
3. Train and test: Once content is selected, the agent will take a little time to index that info. Then you can test it by asking sample questions. See how it responds. If you notice it misunderstood something or gave a subpar answer, you might tweak your content or add specific phrasing it should look for. This is an iterative process initially. You’re teaching your AI.
4. Go live and monitor: Turn it on for real customers! Monitor its performance in HubSpot’s dashboard (you’ll see metrics like number of questions answered by AI, deflection rate (how many conversations didn’t need a human), and so on). Importantly, review transcripts periodically. This will show you if customers are asking things the AI couldn’t handle, which is great feedback to improve either your knowledge base content or the AI’s training. Over time, you should see common questions being handled entirely by the bot, and only the complex issues reaching your team.

One thing to keep in mind (and I tell all my clients this): the AI is only as good as your knowledge base. If your documentation is outdated or missing, the Customer Agent can’t magically invent accurate answers. Make it part of your routine to keep content updated. And that leads us to the fifth and final Breeze agent, which actually helps with this very problem: maintaining and expanding your knowledge base.

Breeze Knowledge Base Agent: Automatically Turn FAQs into Help Articles

The newest member of the Breeze crew is the Knowledge Base Agent, and it’s the perfect complement to the Customer Agent. If you have a growing SaaS product, your knowledge base (KB) articles and FAQs need constant updates and additions. But creating those help articles often gets neglected. Support is busy answering tickets and product is busy shipping features. I’ve seen companies where the knowledge base lags months behind the product, leaving support and customers in the lurch. The Knowledge Base Agent is like a proactive content writer that scours your support interactions for gaps and then drafts new knowledge base articles to fill those gaps. It ensures your self-service documentation stays as dynamic as your product.

What the Knowledge Base Agent Does

In plain terms, this agent learns from all your customer conversations (tickets, chats, emails) and identifies common questions or problems that lack good help articles. Then it goes ahead and drafts knowledge base articles for those topics, so your team can review and publish them. It works hand-in-hand with the Customer Agent: as the Customer Agent encounters questions it can’t confidently answer, the Knowledge Base Agent is alerted: “Hey, we might need an article on this.” Some key capabilities:

• Finds knowledge gaps: For example, if many customers ask “How do I integrate with XYZ?” and you don’t have a help article on that, the agent flags it. It’s continuously mining the language and patterns in support tickets to see what questions are trending or going unanswered in your help center.
• Extracts answers from unstructured data: This is very cool. The agent will sift through support tickets, call transcripts, or email threads related to that question and pull out the relevant information (e.g. how the support rep answered it). Essentially, it gathers the raw material for an article from what your team has already been telling customers informally.
• Drafts a new KB article: Using that info, it writes up a structured article, complete with a suggested title, body content, and maybe bullet points or step-by-step instructions if applicable. It does the heavy lifting of writing in a clear, helpful tone. It also cites where it got the info (so you know it’s based on actual solved cases, which helps with accuracy review).
• Suggests for review: The draft article is then presented to your team. A support lead or knowledge base manager can review the draft, make any edits or additions, and then publish it to your knowledge base. Now, the next time someone asks that question, the answer is documented and easy to find, and your Customer Agent will be able to answer it confidently, closing the loop.

Think of it as turning every resolved support ticket into a learning that improves your documentation. Over time, your KB grows and stays up-to-date without requiring a full-time content writer to comb through tickets manually. More self-service answers means fewer basic tickets, which means your support team can focus on complex issues or proactive outreach (and likely means customers find answers faster on their own).

Why This Matters for SaaS Teams

SaaS customers love good self-service. Many users would rather quickly search your help center than wait on a chat response, especially for non-urgent “how-to” questions. By expanding your knowledge base with relevant articles, you’re improving customer experience and reducing incoming support volume. It’s a direct time saver and cost saver. This agent essentially helps you scale your support content as your product and user base scale, which is crucial for maintaining quality support.

From a marketing perspective, a robust, up-to-date knowledge base can also be a sales asset. Prospects often browse help centers to gauge how well a company supports its product. Showing a rich library of articles (and even public Q&A forums) can build trust. Internally, I’ve seen support teams feel empowered by this agent because it takes the burden off them to document every little thing. The AI has their back, capturing their knowledge as they go about their day.

Another hidden benefit: the Knowledge Base Agent can surface insights to your product team. If the AI finds that “How do I accomplish [a confusing task]?” keeps coming up, that’s a hint the product could be improved or that feature made more intuitive. So you’re not just patching the symptom (lack of an article); you’re informing the cure (maybe fix the UX or add an in-app tip). It’s a beautiful example of AI enabling a feedback loop between support and product.

Setting Up the Knowledge Base Agent

As of writing, the Knowledge Base Agent was in private beta, but it’s expected to roll out broadly to Service Hub users (likely Professional/Enterprise, since it ties into those features). If you have access, here’s what the setup might look like:

• Activate the Agent: Go to Breeze > Agents in HubSpot and see if Knowledge Base Agent is available to turn on. (If not, you might need to request beta access or wait for the general release.)
• Ensure Customer Agent is active: This agent works best alongside the Customer Agent since it uses those interactions to learn. Make sure you’ve set up the Customer Agent to handle support queries; the KB agent will draw from any unanswered questions there.
• Select sources to analyze: You’ll likely specify which past data to learn from. For example, you might let it analyze the last 6 months of support tickets, or all closed tickets tagged “product question”. You might also allow it to read chat transcripts or emails. Don’t worry, this data stays in your system; the AI is just reading it to find patterns.
• Focus on topics (optional): Some setups might let you guide it by telling it to focus on certain areas, e.g. “integrations” or “billing” if you know those are pain points. This can help it zero in faster. Otherwise, it will identify topics on its own.
• Review suggestions: After initial analysis, the agent may present a list of suggested articles or improvements needed. For each suggestion, you can click a “Generate Draft” button. It will then produce a draft article for that topic.
• Edit & publish: Just like any knowledge base article, you’ll review the AI’s draft. Check for accuracy, completeness, and tone. Make edits as needed and publish. Now that topic is covered in your help center! In HubSpot, you might also be able to provide feedback like “this draft was useful” or “needed changes,” which helps refine future outputs.

Since this is a newer tool, keep an eye on HubSpot’s product updates or ask your HubSpot rep for the latest instructions. HubSpot’s Spring 2025 spotlight noted that Knowledge Base Agent helps businesses “instantly enhance and expand their Knowledge Base using existing tickets and conversations.” If you’re on Service Hub Pro or Enterprise, this could be the secret sauce to keeping your help center first-class without massive effort.

Comparison of HubSpot Breeze Agents

To recap, here’s a quick comparison of the five Breeze Agents we’ve discussed, including their key benefits and which HubSpot tier you need to use them:

‍

(All Breeze Agents are part of HubSpot’s AI suite “Breeze” and require the Professional or Enterprise tier of the respective HubSpot Hub. Breeze Copilot and other AI features may be available on lower tiers, but the Agents themselves are premium capabilities.)

This comparison shows how each agent has its niche, but together they cover a wide range of RevOps needs from attracting leads with content, to engaging them on social, converting them in sales, to supporting them post-sale and feeding back into content. Now, how do you put them to work?

Next Steps: Putting Breeze Agents to Work for You

In my journey from marketing manager to consultant, I’ve seen a lot of tools promise to “revolutionize” work. I approach new tech with healthy skepticism because at the end of the day, outcomes matter more than flashy features. HubSpot’s Breeze Agents, however, have earned my trust by delivering tangible results for SaaS teams I work with. They’re not magic wands that remove humans from the equation. Think of them instead as junior teammates who work tirelessly in the background, handling the busywork and giving you the intel you need, so you can focus on strategic, high-impact tasks.

If you’re a SaaS marketer, sales leader, or support lead looking to boost growth and efficiency, here are a few tips to consider as you take next steps with Breeze Agents:

• Check Your HubSpot Plan: Breeze Agents are available on HubSpot’s Professional and Enterprise plans. If you’re already on one of those, you likely have access. Log in and look for “Breeze” in your HubSpot navigation. If you’re on a lower tier (Starter/Free) or not using HubSpot yet, this might be a compelling reason to consider an upgrade or a trial. (HubSpot often offers trials or demos for these AI features – don’t hesitate to reach out to your rep or a solutions partner to explore this.)
• Start Small, Then Expand: You don’t have to turn on all five agents at once. In fact, it might be wise to start with one that addresses your biggest pain point. Maybe begin with the Content Agent to jumpstart your content pipeline, or activate the Customer Agent on a few common FAQs to test the waters. Build confidence and internal buy-in with one success, then roll out more. Each agent will provide value on its own, and together they amplify each other (for example, the Knowledge Base Agent will make the Customer Agent even more effective over time by adding articles).
• Bring Your Team Onboard: Introduce these AI agents to your team in a positive, collaborative way. There can be initial fear (“Is the AI going to take my job?”). It’s crucial to frame it as helping them do their jobs better. Show your content writer how the AI can take care of first drafts (so she can spend more time on creativity and strategy), or show your support rep how repetitive tickets will go down (so he can focus on complex issues or customer success initiatives). When the team sees the agents as helpers, not competitors, you’ll get the best outcomes and adoption. Maybe even let them name the AI agent (like our “Brad”) to build camaraderie!
• Measure the Impact: Before and after you deploy an agent, track key metrics. HubSpot provides dashboards (e.g. number of posts generated, emails sent, tickets deflected). Take note of baseline numbers and then the lift after using Breeze Agents. For example, measure content output (or organic traffic) before vs. after the Content Agent, or average first response time in support before vs. after the Customer Agent. In my experience, the numbers will speak loudly, showing significant time saved or output gained. These data points will help you prove ROI to any stakeholders and justify further investment in AI tools.

At the end of the day, Breeze Agents have made previously impossible workloads “impossibly easy” (to borrow HubSpot’s phrasing). They let lean teams punch above their weight, achieving big-company results with a small-company team. I’ve been continually impressed – and even a little amused – at how I can delegate grunt work to “robots” and get back hours in my day. It’s a reminder that working smarter beats just working harder.

Ready to give your marketing, sales, and support a boost? If you have HubSpot Professional or Enterprise, log in and head to the Breeze section to launch your first agent. If you’re not on those tiers yet, consider upgrading or starting a trial to unlock these features. As a consultant, my advice is simple: don’t leave these AI tools on the table. The SaaS world moves fast, and those who leverage AI helpers like Breeze Agents will have an edge in growth and efficiency. Let the agents handle the busywork and data-crunching, so you can focus on strategy, creativity, and building customer relationships. The things humans do best.

Give Breeze Agents a try, and I suspect you’ll soon wonder how you managed without your new digital coworkers. Happy scaling!