Service response
Retrieves approved knowledge and account context, then answers, requests what is missing, or routes to the right person.
AI for Customer Success
AI customer service, built inside the CRM you already run
The fear is not that AI answers slowly. It is that it answers wrongly, to a customer, in writing. So we start by deciding what it is allowed to say.
.png?height=28&name=Gunnebo%20Logo%20(1).png){kind=small}
.png?height=28&name=Closer%20Still%20Media%20(1).png){kind=small}
.png?height=28&name=Ecommpay%20Logo%20(1).png){kind=small}
.webp?height=28&name=Tussell%20Logo%20(1).webp){kind=small}
-4npmo6jxod.png?height=28&name=iconsolutionsrgb%20(1)-4npmo6jxod.png){kind=small}
What we usually find, and what we build instead
Four things we find in almost every service team that has tried a bolt-on chatbot.
Without SpotDev
What we usually find
With SpotDev
What we build instead
Without SpotDev
The same twenty questions, every week
Skilled people answering what a published article already answers, while real problems queue behind them.
With SpotDev
Published answers handled
Routine questions resolved or drafted, so complex work reaches a person sooner.
Without SpotDev
The answer is in a system the chat widget cannot see
Order status, contract terms and delivery dates sit in the ERP. The bot sees the help centre.
With SpotDev
Retrieval from the real source
Answers grounded in the systems that actually hold the fact, not just the knowledge base.
Without SpotDev
You find out about churn at renewal
The signals were there for months, spread across tickets, usage and unanswered emails.
With SpotDev
Risk surfaced with its evidence
Conditions detected against CRM evidence, with the reasoning shown so a human can judge it.
Without SpotDev
Nobody can say what the AI may say to a customer
So it is either switched off, or switched on and quietly frightening.
With SpotDev
A written action list per workflow
Granted one action at a time, with everything sensitive escalated from day one.
What a customer success wave activates
A wave activates two of the four agent capabilities, chosen against your own value case. Enablement and measurement are part of every wave, not alternatives to them.
Account review preparation
Assembles outcomes, risks, open requests and commitments before the review, so the meeting starts with the facts already gathered.
Retention risk detection
Detects inactivity, unresolved service and adoption decline, explains the evidence, and recommends an intervention for a person to take.
Account growth signals
Connects a request or a changed condition to a credible expansion hypothesis, and prepares the internal briefing behind it.
Enablement for the service team
The people answering customers trained on when to trust a draft, when to correct it, and how to report one that was wrong.
Measurement against a baseline
Response time, resolution time, escalation and correction rate, tracked against the position we recorded before anything was built.
Where HubSpot's own agent stops
- 01
It does more than answer articles
HubSpot's knowledge base says the customer agent performs configured actions, giving password resets and order-status checks as examples. Treat it as capable, then find the edge.
- 02
Every action is configured one at a time
Each action is set up individually against an external app. That is fine for a handful. It is not a shared context layer your whole estate can draw on.
- 03
Sensitive actions stay at arm's length
HubSpot's own guidance is that anything touching account access or credentials should send a secure link rather than make the change in the conversation.
- 04
Three of the four jobs are not deflection
Review preparation, retention risk and account growth are not ticket answering, and nothing built for deflection is built for them.
What it is allowed to say
Authority is granted per action, not per agent. Check what your own subscription covers, then decide each of these deliberately.
Always escalated
Refunds, complaints, legal matters, account closure and anything about a vulnerable customer.
External send is separate
Permission to draft is not permission to send. The two are never granted together by default.
Confidence is not a control
A model's confidence score is not evidence. We gate on source and action type instead.
Fails safely
When identity, entitlement or the source is uncertain, it stops and hands over rather than guessing.
Often permanent draft
For many customer-facing workflows, drafting for a human is the right end state, not a phase.
Everything is logged
Every answer, escalation and correction recorded, so a complaint can be reconstructed from evidence.
Implementation expertise that drives adoption
An in-house engineering team that has delivered more than 300 technology projects. Work with engineers who have built this before.
300+HubSpot implementations, firm-wide
In-house engineering team.
Agents and integrations designed, tested and shipped by our own engineers. Nothing is subcontracted.
OpenAI Select Partner and Claude Registered Partner
SpotDev is an OpenAI Select Partner and a Claude Registered Partner, alongside HubSpot Diamond status and Custom Integration and Onboarding accreditations.
Model chosen for the job.
Honest evaluation of the right model for the work. We will say when the answer is the AI you already pay for.
AI with security in mind
- Not trained on by default
- Cyber Essentials Plus
- Built on your infrastructure
- In-house engineers
Common questions
What happens when the AI gets it wrong?
It should mostly not be in a position to. Anything consequential is drafted for a person. Where it does answer directly, errors are logged, reviewed and fed back into the evaluation set.
Will our customers know they are dealing with AI?
We recommend telling them, because trust is easier to keep than rebuild. That is a design choice we advise, not a legal opinion, and your own advisers should confirm your obligations.
Is our customer data used to train a model?
OpenAI states on its enterprise privacy page, updated 8 January 2026, that it does not train on business data by default unless you explicitly opt in. We configure for that and verify it.
We already have HubSpot Service Hub. Is that enough?
Often, yes, and we will say so. It is enough when the answers genuinely live in your published content. It stops when the answer lives in another system.
Do we need to buy another platform?
No. That is the argument of this page. Everything else in this market sells you a chatbot platform to run alongside your CRM, which is how you end up with two versions of the customer.
Our knowledge base is out of date. Does that stop us?
It changes the first job. Grounding an agent in stale content produces confident wrong answers, so content quality gets assessed before anything faces a customer.
What does it cost?
From £10,000 per month as part of AI Accelerator, committed a quarter at a time. Model and platform usage is billed separately. Enterprise programmes are priced on complexity.
Start with a diagnostic.
A short, fixed-scope assessment of which service workflows are safe to automate, which are not, and what your content and data can currently support. No obligation.
SpotDev is an OpenAI Select Partner, a Claude Registered Partner and a HubSpot Diamond Solutions Partner. SpotDev is a separate company from OpenAI and from Anthropic, and partner status does not constitute endorsement of SpotDev by OpenAI or by Anthropic. OpenAI and ChatGPT are trademarks of OpenAI. Claude is a trademark of Anthropic, PBC. HubSpot is a trademark of HubSpot, Inc.