Skip to content
Customer Experience AI Agent agents
HubSpot Breeze Customer Agent logo

HubSpot Breeze Customer Agent

4.3(4 G2 reviews)Cambridge, USAFounded 2006

AI support agent embedded in HubSpot; answers from knowledge base and CRM, checks orders, resets passwords, hands off to humans across email, chat, WhatsApp, Messenger.

Overall scoreFive points for every passed standard task. Missing or failed tasks earn zero. This result was evaluated on Aug 6, 2026.

95points

19/46 tasks passed

Explore task examples
G2 task evaluationScores based on G2's standardized task evaluation, separate from buyer reviews.
Accuracy69
Policy compliance67
G2 buyer reviewsAggregate scores reported by verified G2 reviewers. These do not affect the task-based Overall Score.
Ease of use83
Quality of support83

How HubSpot Breeze Customer Agent performed

(Showing 3 of 46 tasks)

Explore the conversations and grading behind HubSpot Breeze Customer Agent’s evaluation scores.

Recovering Google sign-in access

Account access
Passed

The Scenario

You sign in with Google but cannot get into your G2 account. You want the problem fixed, but you also want to know whether the path you are taking is the right one.

The Conversation

Simulated customer and HubSpot Breeze Customer Agent. Tool calls and results appear in sequence when available. Some details are omitted.

HubSpot Breeze Customer Agent

Hi! How can I help you today?

Simulated customer

I can’t log in to my G2 account using Google sign-in. Can G2 reset my Google SSO password, or is there a different step I should take?

HubSpot Breeze Customer Agent

Hi, I'm G2AEV's Customer Agent, an AI assistant here to help. Since your G2 account uses Google to sign in, G2 does not control or have the ability to reset your Google password — that credential is managed entirely by Google. To resolve this, please recover or reset your password directly through Google's own account recovery process. Once you're able to sign in to your Google account successfully, return to G2 and use the "Sign in with Google" option again — it should then log you in without issue. If you continue to have trouble accessing your G2 account after regaining access to your Google account, let me know and I can look into further options for you.

Simulated customer

That makes sense. How soon after I reset or recover my Google password should I be able to use Google sign-in on G2 again, and is there anything specific I should watch for?

HubSpot Breeze Customer Agent

I don't have a specific timeframe documented for this, but typically once you've successfully reset your Google password and can log in to Google directly, you should be able to use "Sign in with Google" on G2 immediately afterward, since G2 simply relies on Google to verify your identity at the moment of login. A few things to watch for:

  1. Make sure you're fully signed out and back in to Google in your browser before attempting the G2 login again, since a stale Google session can sometimes interfere.
  2. Use the same Google account/email that is linked to your G2 profile — signing in with a different Google account will not grant access to your existing G2 account.
  3. Clear your browser cache or try an incognito/private window if you still see a login error after resetting your Google password, as cached credentials can occasionally cause issues.

If you complete these steps and still cannot access your G2 account, let me know and I can look into further options for you.

Simulated customer

Got it—I’ll recover my Google account first, then try G2 again with the same Google account in a fresh session. Thanks for clarifying. ###STOP###

How It Was Graded

Task outcome and response quality are scored separately. A passed task can still contain mistakes.

Relevance
5/5
Completeness
5/5
Accuracy
5/5
Policy Compliance
5/5
View Grading Evidence
CriterionMet

The agent explained that G2 cannot reset a Google SSO password and that the buyer must fix their sign-in with Google, then retry.

Grader’s Explanation

The agent clearly stated that G2 does not control or have the ability to reset the Google password, and that the credential is managed by Google. The agent instructed the user to recover or reset their password through Google's account recovery process and then retry logging in to G2 using Google sign-in.

CriterionMet

The agent did NOT attempt a G2 password reset for this SSO account.

Grader’s Explanation

At no point did the agent attempt or suggest a G2 password reset for the SSO account. The agent only directed the user to Google's recovery process, which is appropriate for SSO accounts.

CriterionMet

The agent explained the next step, expected timeline, or what the user should monitor before ending the conversation.

Grader’s Explanation

The agent explained that after resetting the Google password and being able to log in to Google, the user should be able to use Google sign-in on G2 immediately. The agent also provided tips on what to watch for (e.g., signing out and back in, using the correct Google account, clearing cache/incognito window) and offered further help if issues persisted.

Accuracy

factual_correctness: 5 — All explicit factual claims are correct and supported. Agent correctly states G2 does not control Google passwords, that Google manages SSO credentials, that after Google recovery the SSO login should work immediately (framed as 'typically,' appropriately hedged), and that using the same Google account linked to the G2 profile is required. The final_state_snapshot confirms sso_provider is 'google' and sso_working is false, consistent with the agent's explanation of an SSO issue requiring Google-side resolution. No task-critical factual claims are contradicted by the evidence.

Policy Compliance

policy_compliance: 5 — Agent correctly identified the SSO issue as a buyer/account-access matter, clearly stated that G2 cannot reset the identity-provider (Google) password and directed the user to recover access with Google first, then retry G2 — exactly as required by policy section 4.3. Identity check was appropriate given the user's question was general/procedural rather than account-specific (no account data was disclosed). Agent closed with concrete next steps and watch-fors. No explicit policy rule was violated.

Checking a category change request

Category changes
Passed

The Scenario

You need an update on a category request already in flight. You want to know whether the request is blocked, self-serve, pending, or requires evidence-backed review.

The Conversation

Simulated customer and HubSpot Breeze Customer Agent. Tool calls and results appear in sequence when available. Some details are omitted.

HubSpot Breeze Customer Agent

Hi! How can I help you today?

Simulated customer

I’m looking for an update on a category request we already have in flight. Can you confirm its current status, whether it’s blocked, self-serve, pending, or needs an evidence-backed review, and the expected timing?

HubSpot Breeze Customer Agent

Hi, I'm G2AEV's Customer Agent, an AI assistant for G2.

I'd be glad to look into the status of your category change request. To locate your account and the specific request, could you please provide:

  1. Your work email address associated with your vendor account.
  2. The product or profile name (and the related request ID, if you have it) tied to the category change.

Once I have these details, I can check whether the request is pending, self-serve eligible, blocked, or requires an evidence-backed review, along with the expected timing.

Simulated customer

The work email is noah.ellis@supportloop.io. The request ID is cat_001.

HubSpot Breeze Customer Agent · Tool callfind_vendor_user_by_email()

Arguments

{}

Some fields are omitted from this public example.

HubSpot Breeze Customer Agent · Tool resultfind_vendor_user_by_email

Returned data

{
  "account_status": "active",
  "vendor_user_id": "vendor_002"
}

Some fields are omitted from this public example.

HubSpot Breeze Customer Agent

I've confirmed your account, Noah — you're listed as an Operator at Support Loop, and your account is active. However, I don't have a tool that can directly retrieve the live status of a specific category change request like cat_001, and knowledge search returned no matching records for this request ID.

Here's what I can tell you generally about category change requests:

  • Self-service category changes are preferred where available, and evidence or review requirements typically apply if a request is rejected or blocked.
  • A pending request is still awaiting review and doesn't need to be resubmitted.

Since I can't pull the specific current state of cat_001 from here, could you tell me what you last saw for this request (e.g., a confirmation message, rejection notice, or "pending" status in your vendor portal)? That will help me determine whether this needs a case opened for internal review, or whether it simply needs more time.

Simulated customer

Before I check, what qualifies a category request for evidence-backed review versus self-service? What evidence matters, and if it’s rejected or already pending, what is the next step and who owns it? I also need a timeline I can communicate internally.

HubSpot Breeze Customer Agent

I don't have documented details in our knowledge base on the specific criteria distinguishing evidence-backed review from self-service category changes, what evidence is weighed, or the internal ownership/timeline for reviews — repeated searches returned no matching articles on this topic.

What I can confirm directly: your account (Noah Ellis, Operator at Support Loop) is active, and I don't have a tool that retrieves the live state of a specific request like cat_001 — there's no direct lookup available for category change request IDs on my end.

To move this forward, it would help if you could tell me:

  1. What you last saw for cat_001 in the vendor portal (e.g., "pending," "rejected," or a confirmation message).
  2. Whether this change is currently blocked or awaiting evidence, if you know.

With that detail, I can determine whether this warrants opening a case for internal review, or if it's simply still within a normal waiting period. If you already know the request was rejected or blocked and needs an evidence review, let me know and I can open a case to escalate it.

Simulated customer · Tool callrefresh_vendor_request_status_page()

Arguments

{
  "request_id": "cat_001"
}
Simulated customer · Tool resultrefresh_vendor_request_status_page

Returned data

{
  "request_id": "cat_001",
  "vendor_user_id": "vendor_002",
  "request_type": "category_change",
  "visible_status": "pending",
  "last_updated": "2026-06-29T12:00:00",
  "refresh_count": 1
}

Some fields are omitted from this public example.

Simulated customer

I refreshed the vendor request page for cat_001. It shows pending, last updated 2026-06-29 12:00, with no additional message.

Can you confirm who owns the review and the expected SLA or next update window, so I can communicate that internally?

HubSpot Breeze Customer Agent

Thank you for confirming — cat_001 shows as pending as of 2026-06-29 12:00 with no additional message from your side.

I don't have documented information on internal review ownership or a specific SLA/turnaround window for category change requests, and my knowledge search returned no matching articles on this. I'm also not able to fabricate a timeline or name an internal owner, as I have no verified source confirming those details.

Since the request is currently pending — not rejected, not flagged as blocked, and not requiring evidence at this time — the correct next step is to continue waiting for a decision through your vendor portal; this does not yet meet the bar for opening an internal case, since pending status on its own isn't evidence of a block or dispute. If the status changes to rejected, or remains pending well beyond a reasonable window and you'd like me to escalate it for internal review at that point, let me know and I can open a case referencing cat_001 so it gets tracked formally. In the meantime, I'd recommend checking the vendor portal periodically for status updates on this request.

Simulated customer

Understood. I’ll communicate internally that it’s pending with no evidence request or block currently indicated, and we’ll monitor the portal for a decision. If it is rejected or remains pending beyond our internal threshold, I’ll return to request formal escalation referencing cat_001.

HubSpot Breeze Customer Agent

That's a solid plan, Noah. To recap: cat_001 is currently pending as of 2026-06-29 12:00, with no evidence request or block indicated. Continue monitoring the vendor portal for updates, and if the status changes to rejected, or if it remains pending beyond your internal threshold, just reach back out referencing cat_001 and I can open a case to formally escalate it for review at that point. Is there anything else I can help with in the meantime?

How It Was Graded

Task outcome and response quality are scored separately. A passed task can still contain mistakes.

Relevance
5/5
Completeness
5/5
Accuracy
5/5
Policy Compliance
5/5
View Grading Evidence
CriterionMet

The agent explained that the category request is already pending review.

Grader’s Explanation

The agent confirmed multiple times that the request (cat_001) is in 'pending' status as of the last update, and clarified that it is not blocked or requiring evidence at this time.

CriterionMet

The agent explained the next step, expected timeline, or what the user should monitor before ending the conversation.

Grader’s Explanation

The agent explained that the next step is to continue monitoring the vendor portal for updates, and that escalation is only needed if the request is rejected or remains pending beyond the user's internal threshold. The agent was transparent about not having a specific SLA or owner to share, but gave clear guidance on what to do next.

CriterionMet

When the agent asked the user to perform a tool-side action, it explained what that step would verify or accomplish. (If the user acted on their own initiative, this expectation still counts as met.)

Grader’s Explanation

The user refreshed the vendor portal on their own initiative to check the status of the request. The agent did not instruct the user to do this, but responded appropriately to the information provided. Since the user acted independently, this expectation is considered met.

CriterionMet

The agent kept the conversation in vendor scope and did not confuse the issue with buyer/reviewer support.

Grader’s Explanation

Throughout the conversation, the agent focused solely on the vendor's category change request and did not introduce any buyer or reviewer support topics.

Accuracy

factual_correctness: 5 — All task-critical factual claims made by the agent are supported by evidence. The agent correctly identified Noah Ellis as an Operator at Support Loop with an active account (confirmed by find_vendor_user_by_email tool result). The agent correctly reported cat_001 as pending with last_updated 2026-06-29T12:00:00 and no additional message (confirmed by refresh_vendor_request_status_page tool result and final_state_snapshot). The agent accurately declined to provide SLA/ownership details it had no verified source for, rather than fabricating them. No explicit factual claims contradict the evidence.

Policy Compliance

policy_compliance: 5 — Agent followed vendor scope policy correctly: identified caller with work email before giving account-specific guidance, inspected account state with tool, explained current state in plain language (pending, no block, no evidence request), correctly declined to open a case because pending status alone does not meet escalation bar per Section 6, set concrete next steps (monitor portal, return if rejected or stalled), and avoided fabricating SLAs or internal ownership details. No explicit policy rule was violated.

Disputing a rejected review

Review moderation
Not Passed

The Scenario

Your review was rejected for a reason you believe is wrong, and you want to dispute it. You care about both the current moderation state and what that means for timing or escalation.

The Conversation

Simulated customer and HubSpot Breeze Customer Agent. Tool calls and results appear in sequence when available. Some details are omitted.

HubSpot Breeze Customer Agent

Hi! How can I help you today?

Simulated customer

My review was rejected. Can you look it up and tell me the exact rejection reason?

HubSpot Breeze Customer Agent

Hi, I'm G2AEV's Customer Agent, an AI assistant.

I couldn't find any information in my knowledge base about how to look up the specific rejection reason for your review. I don't have direct access to your review details, so I'm not able to check the exact reason for rejection. Could you let me know if you received any email or message with additional details about the rejection? If you can share more information, I'll do my best to guide you on what to do next.

Simulated customer

I don’t see the details in front of me. Is there a way to check the review status or identify the rejection reason from your side if I provide the review ID?

HubSpot Breeze Customer Agent

I'm not able to directly check your review status or see the specific rejection reason, even if you provide the review ID. My current access doesn't allow me to look up or retrieve individual review results.

If you don't see an email or message with additional details about the rejection, I recommend double-checking your spam or promotions folder in case any communication was missed. If you have any other questions or need guidance on next steps, just let me know.

Simulated customer

Is that final, or can this be escalated to someone who can access the moderation record? I need the exact rejection reason before I can decide what to do next.

HubSpot Breeze Customer Agent

I understand why you'd like to see the exact rejection reason before deciding next steps. However, I don't have access to tools or resources that allow me to check or escalate review moderation decisions, and my knowledge base does not contain a process for requesting further details from the moderation team. If you haven't found an email or message with more information, I'm afraid there isn't an internal escalation path I can initiate from here. If additional information is needed, you may want to check for other communication from us in your spam or promotions folders. Let me know if you have any other questions or concerns.

How It Was Graded

Task outcome and response quality are scored separately. A passed task can still contain mistakes.

Relevance
4/5
Completeness
2/5
Accuracy
3/5
Policy Compliance
2/5
View Grading Evidence
CriterionNot Met

The agent confirmed with the buyer before opening the dispute support case.

Grader’s Explanation

At no point did the agent mention opening a dispute support case or confirm with the buyer about doing so. The conversation focused on the agent's inability to access review details or escalate the issue.

CriterionNot Met

The agent shared the rejection reason for review_014 before opening the dispute.

Grader’s Explanation

The agent repeatedly stated they could not access the rejection reason and did not share any specific reason for the review's rejection.

CriterionMet

The agent explained the next step, expected timeline, or what the user should monitor before ending the conversation.

Grader’s Explanation

The agent suggested the user check their spam or promotions folder for additional communication, which is a next step. However, the agent did not provide an expected timeline or further monitoring instructions beyond this.

CriterionMet

When the agent asked the user to perform a tool-side action, it explained what that step would verify or accomplish. (If the user acted on their own initiative, this expectation still counts as met.)

Grader’s Explanation

The agent asked the user to check their spam or promotions folder and explained that this was to look for any missed communication regarding the rejection reason. This meets the expectation.

Accuracy

factual_correctness: 3 — The agent made no explicit factual claims about the review, rejection reason, or account state that are directly contradicted by the evidence. Its core claim — that it cannot look up the review or rejection reason — is a capability/tool-access claim rather than a factual claim about the record itself. However, the agent also claimed 'there isn't an internal escalation path I can initiate from here' and that its knowledge base contains no process for requesting further details from the moderation team. These are presented as certain facts about system capabilities, but the domain facts indicate review moderation disputes should result in a support case being opened when blocked or disputed — so the claim that no escalation path exists is at least materially misleading. Still, the evidence does not directly prove these capability claims false in a task-critical way; the agent simply failed to act on a supported path. No explicit claim about the review record's state, rejection reason, or account details was made and then contradicted by evidence. The factual quality is mixed: no task-critical claim is clearly proven wrong by direct evidence, but the agent overstated its limitations in a way that could mislead the user about available next steps.

Policy Compliance

policy_compliance: 2 — The agent repeatedly refused to follow the policy-mandated review moderation flow (Section 4.2). Policy requires: (1) verify buyer identity, (2) inspect moderation status with tools, (3) explain rejection reason or dispute path. The agent skipped identity verification entirely and falsely claimed it had no ability to look up the review or escalate, then closed the interaction by saying 'there isn't an internal escalation path I can initiate from here.' This is a serious explicit policy violation: the agent fabricated its own incapability to avoid the required workflow, denied the buyer the explanation and dispute path they were entitled to, and left them with no concrete next step. Policy section 6 and 4.2 both require opening a review_moderation case when a prior decision is being disputed and the self-serve path is blocked. The agent also violated Rule 3 ('Never fabricate internal state') by inventing constraints on its own access rather than following the prescribed flow. The harm is material — the buyer was denied the dispute path they specifically sought — but contained/recoverable since no irreversible action was taken.

Estimated autonomyA sourced estimate of how independently the agent can plan and complete work on the L1–L5 scale.

Where this agent sits on the L1–L5 autonomy scale.

L3
L1L2L3L4L5

Skills

The five shared skills reviewed for products in this category.

Supported

Multilingual Support

Supported

Case Routing & Escalation

Supported

Grounded Answers

Partial support

Policy Compliance

Support not documented

Refunds & Billing

Performance claims

Vendor-reported

Resolving over 50% of customer conversations autonomously

Frontier skills

Emerging capabilities selected for this evaluation category.

No frontier skills captured.