CRM AI AGENTS · HUBSPOT · SALESFORCE

AI agents that qualify leads and update records in HubSpot and Salesforce

Your reps spend the end of each day logging calls, fixing fields and chasing stale deals. We build agents that do that work inside HubSpot or Salesforce: they score inbound leads, move deal stages, log activity, fill missing contact data and draft follow-ups. Your rules decide which actions need a person to approve them.

What brings you here?
Working software on your own data before you sign. Bhavesh replies within 24 hours.

Have a brief or RFP? Attach it at step 2. We scope, price and sign each phase on its own.

500+ businesses served97% on-time deliveryFixed price per phaseYou own the code

ONE LEAD, FOUR ACTIONS

A demo request arrives at 9:02. The CRM is up to date by 9:03.

  1. 1

    Contact created from the form. Company size and industry filled from the website, source logged.

  2. 2

    Scored against your ideal customer rules. Owner set, reason written on the record.

  3. 3

    Follow-up email drafted in the owner's name. Waiting for the rep to send.

  4. 4

    Deal created at Discovery. Stage changes later need the owner's one-click approval.

Illustration of a typical lead flow. Your rules and fields set the real steps.

THE SHORT ANSWER

What does an AI agent do inside a CRM?

An AI agent for HubSpot or Salesforce reads your CRM records, decides on a next step and makes the change through the CRM's own API. It can qualify leads, update deal stages, log calls and emails, enrich contacts and draft follow-ups. Your rules decide which actions need a person's approval first.

FactoryJet builds custom AI agents for HubSpot and Salesforce that qualify leads, update deal stages, log calls and emails and enrich contacts, with a demo on your own records before you sign. AI agents for CRM are a type of agentic AI, meaning AI that takes actions in your systems after reading them.

Updated 31 October 2026 · Written by Bhavesh Barot, Founder & CEO

WHO THIS IS FOR

  • Sales leaders and RevOps (revenue operations) leads who run a CRM for a team of reps.
  • Companies on HubSpot Sales Hub, Salesforce Sales Cloud, or both.
  • Teams where CRM data goes stale because logging comes last.
  • Buyers with a written brief or RFP who want a fixed price.

BY THE NUMBERS

  • 500+

    businesses served

  • 97%

    of projects delivered on time

  • 2014

    the year FactoryJet started building

WHAT WE BUILD

Six CRM jobs an AI agent can take on

Start with one. Each job is scoped and priced as its own phase, so you can prove it on your data before adding the next.

1. Qualify inbound leads

A form fill or chat creates a contact. The agent reads the answers, the company website and past activity, scores the lead against rules you write in plain English, sets the owner and status, and writes the reason next to the score.

HubSpot contacts or Salesforce leads

2. Update deal stages

The agent reads replies, call notes and meeting outcomes, then proposes the next stage with the evidence attached. The deal owner approves with one click. Closed Won always stays with a person.

Deals or opportunities

3. Log calls and emails

Each email thread, call summary and meeting note is matched to the right contact, company and deal, then logged as an activity with next steps. When the match is unclear, it goes to a review list. The agent does not guess.

Calls, emails, meetings, notes

4. Enrich contacts and companies

Missing job titles, company size, industry and website are filled from sources you approve, such as the contact's own emails, the company site or a data provider you already pay for. Each value records where it came from.

Contact and company properties

5. Draft follow-ups

After a call or a quiet week, the agent drafts the follow-up from the deal history and the last conversation. The rep edits and sends it. Drafts sit on the record, so a manager can see what went out.

Email drafts and tasks

6. Work across your other systems

Check open invoices in QuickBooks or NetSuite before a renewal call, read Zendesk tickets before a deal review, or sync a lead between HubSpot marketing and Salesforce sales with one shared ID.

ERP, help desk, inbox

HOW IT CONNECTS

How a Salesforce AI agent or HubSpot AI agent connects to your CRM

No new screen for your reps. The agent writes to the records they already open, and every change shows up in the record history.

In HubSpot, the CRM is built from objects, records and properties. A contact is a record of the contact object, an email address is a property, and a contact links to its company and deals through associations. HubSpot explains this model in its guide to the CRM APIs. The agent reads and writes those same objects. HubSpot also runs an MCP server that gives AI tools read and write access to contacts, companies, deals, tickets and activities, under the scopes of an app you approve. MCP (Model Context Protocol) is a standard way for an AI tool to call an app's functions.

In Salesforce, the agent signs in through a connected app, which is an app registered in your org with its own keys. It uses an integration user, a Salesforce login kept only for the agent. A permission set, the list of what that login may see and change, limits it to the objects and fields the job needs, such as Lead, Contact and Opportunity. Switch on field history tracking for the fields the agent edits, and Salesforce shows each change on the record. The agent's own action log also records every change with its reason.

The model itself (Claude, GPT or Gemini) never touches the CRM directly. It uses tool use to ask for an action by name. Our code checks that action against your rules, and only then does the API call run. If the job also needs your ERP or help desk, our AI integration services and our ERP AI agents cover those links, and our guide to a custom MCP server for your ERP or CRM explains when you need one.

Approval rules we set with you

  • Draft a follow-up email

    Always drafted. A rep reviews and sends.

  • Move a deal stage

    Suggested with the reason. The owner approves. Low-risk early stages can run alone later.

  • Fill an empty field

    Runs alone, with the source logged on the record.

  • Change a field a rep typed

    Never, unless your written rules allow it for that field.

  • Merge or delete records

    Not allowed. Duplicates go to a review list.

  • Send an email to a customer

    Off by default. Narrow cases, such as a meeting confirmation, only if you switch them on.

Every action is logged with the old value, the new value and the reason.

HOW A PROJECT RUNS

From your brief to a live CRM agent in five steps

The agent earns trust in stages. It suggests first, and acts alone only where the review numbers from your own CRM support it.

  1. Send the brief

    Tell us your CRM and edition, the first job, how many leads or deals a week, and who approves the agent. Have an RFP? Send it as is.

  2. Demo on your records

    Before you sign, the agent runs on a sample of your real records, with names removed if you prefer, and shows each decision with its reason.

  3. Fixed scope and price

    You get the jobs, the approval rules and one fixed price per phase in writing. Each phase is scoped, priced and signed on its own.

  4. Build in a sandbox

    We build against a HubSpot test account or a Salesforce sandbox, never your live data first, with a weekly demo inside your own CRM.

  5. Launch with suggestions

    The agent starts by suggesting changes. We review accepted, edited and wrong suggestions with your RevOps lead, then loosen rules where the numbers support it.

Next step

Get a CRM agent demo on your own records

Send your brief or RFP. Tell us your CRM, the first job you want off your reps' plates and who approves changes. We reply with what we would build and a date for a demo on your records.

See a demo on my dataBhavesh replies within 24 hours.

HOW WE WORK

What terms do you sign before the build starts?

FactoryJet quotes the build, and any support you choose, as fixed prices. We design the agent, build it, launch it and stay on to support it if you want us to.

  1. 01

    Demo before you sign

    You see the agent working on your own records before any contract.

  2. 02

    You own the code

    The repository, the CRM app and the written rules are yours from day one.

  3. 03

    Your accounts, at cost

    The AI model and hosting can run in your own accounts, billed to you at cost with no markup.

  4. 04

    Fixed price per phase

    Each phase is scoped, priced and signed on its own, so you can stop after the first job.

  5. 05

    Defects fixed in the build price

    Defects found after launch are fixed inside the build price, for a cover period agreed before work starts.

  6. 06

    Support only if you want it

    After that, each change is quoted and approved before we do it. There is no monthly minimum.

YOUR OPTIONS

Custom agent vs CRM vendor agent vs a Zapier workflow

Each one fits a different team. Agentforce is Salesforce's own agent platform, and Breeze is HubSpot's set of AI tools. A vendor agent like these suits work that stays in one CRM. A Zapier or Make workflow suits fixed if-this-then-that steps. A custom agent suits jobs that read free text, cross systems or need your own rules.

Custom agent (FactoryJet)CRM vendor agent (Breeze, Agentforce)Zapier or Make workflow
Reads free text, such as emails and call notes.YesYesPartial
Works across HubSpot and Salesforce in one flow.YesNoYes
Reads your ERP, help desk or inbox before acting.YesPartialPartial
Approval rules written per action and per field.YesPartialPartial
You own the code and can move it.YesNoThe workflow stays in your Zapier or Make account and runs only there.
Pricing model.Fixed price per phase.Vendor licence or credits.Plan tier by task volume.

Swipe the table sideways to see every column.

Vendor details checked on 31 October 2026 against: Salesforce Agentforce developer guide, HubSpot AI (Breeze), Zapier, Make. Vendors change their products often, so check the linked page before you decide.

AGENT WORK WE HAVE SHIPPED

What agent work has FactoryJet already shipped?

For Washington Law Group, a US law firm, we built an AI agent that finds serious crashes involving commercial vehicles in published news and police sources, checks the facts against the article text, merges duplicate reports into one record and emails the firm. Lawyers review every lead in a private console. A CRM agent follows the same pattern: it gathers and checks, and your team decides.

Read the Washington Law Group case study

RELATED PAGES

FREQUENTLY ASKED QUESTIONS

Questions sales and RevOps leaders ask about CRM AI agents

Straight answers on what the agent can change, who approves it, and what you own.

What It Is

Can an AI agent read and update HubSpot or Salesforce records?

Yes. Both CRMs let approved apps read and write records through their APIs. HubSpot organises its CRM into objects such as contacts, companies, deals and tickets, and its MCP server gives AI tools read and write access to them. In Salesforce, the agent signs in through a connected app with its own user and permission set. We decide with you which objects and fields the agent may change. Everything else stays read-only or hidden.

What is a CRM AI agent?

A CRM AI agent is software that reads your CRM, decides on a next step and acts on it. Examples are scoring a new lead, moving a deal stage or drafting a follow-up email. An AI model such as Claude or GPT does the reading and writing. Rules you set decide what the agent may change alone and what a person must approve first.

How is a custom agent different from HubSpot Breeze or Salesforce Agentforce?

Breeze and Agentforce are agents the CRM vendors sell inside their own products. They suit teams whose work stays inside one CRM. A custom agent fits when the job crosses systems, such as reading a quote in your ERP before updating a deal, or when you want your own approval rules and code you own. Our Agentforce pricing guide compares the two on cost and fit.

Is a CRM agent the same as an AI SDR?

They do different jobs. An SDR is a sales development rep, the person who works new leads. An AI SDR works at the edge of the CRM: it writes to new prospects, replies to inbound leads fast and books meetings. A CRM agent works on the records your team already has, keeping them accurate and moving deals forward. Many sales teams run both, and they can share the same lead scoring rules.

Do we need Agentforce to run an AI agent in Salesforce?

No. Agentforce is Salesforce's own agent platform, and you can use it if it fits. A Salesforce AI agent can also be built outside it. The Salesforce REST API lets an approved app create, read, update and delete records, so a custom agent signs in through its own connected app and works on leads, contacts and opportunities. Our Agentforce pricing guide compares the cost of each route.

Do you build agents for Dynamics 365, Zoho or Pipedrive?

This page covers HubSpot and Salesforce. Microsoft Dynamics 365, Zoho CRM and Pipedrive each publish an API for reading and writing records, so an agent connects to them the same way: its own login, limited permissions and your approval rules. Send the CRM name with your brief and we confirm what its API allows before we quote.

Will AI agents replace our CRM or our CRM admin?

No. The agent works inside the CRM you already pay for, and the records stay there. It takes over repeat work such as logging calls, filling empty fields and drafting follow-ups. Your CRM admin still owns the fields, the pipeline stages and the rules the agent follows. Your admin reviews the agent's suggestion log instead of fixing stale records by hand.

What It Does

Which CRM tasks can an AI agent handle?

Six jobs cover most of the requests we get. Qualifying inbound leads against your ideal customer rules. Updating deal stages when an email or call shows progress. Logging calls, meetings and emails to the right record. Enriching contacts and companies with missing fields. Drafting follow-ups for a rep to send. Reading your ERP or help desk before acting. Each one can be built alone, or several can be combined into one agent.

Can the agent qualify inbound leads in HubSpot?

Yes. When a form or chat creates a contact, the agent reads the form answers, the company website and any past activity. It scores the lead against rules you write in plain English, then sets the lead status and owner. It also adds a short note that explains the score, so a rep can read the reason before calling.

Will the agent move deals between stages on its own?

Only if you allow it. Most teams start with suggestions. The agent reads a reply or call notes, proposes a stage change with the reason, and the deal owner approves it in one click. Once suggestions prove right for a few weeks, you can let low-risk moves happen without approval, such as Discovery to Demo Booked. Closed Won stays with a person.

How does contact enrichment work, and where does the data come from?

The agent fills missing fields such as job title, company size, industry and website from sources you approve. Those can be the contact's own emails, the company website, or a data provider you already pay for, such as Apollo or ZoomInfo. It records where each value came from. It never overwrites a field a rep typed by hand unless your rules allow it.

Does the agent send emails to customers by itself?

Not by default. The agent drafts the follow-up inside the CRM, using the deal history and the last conversation, and a rep reviews and sends it. Some teams later allow automatic sends for narrow cases, such as a meeting confirmation. You set those rules in writing before launch, and every email the agent sends is logged on the record.

Can one agent work across HubSpot and Salesforce at the same time?

Yes. Some companies run marketing in HubSpot and sales in Salesforce. The agent can read a lead's marketing activity in HubSpot, then update the matching lead or opportunity in Salesforce, using one shared ID so the records stay matched. Before the build, we agree which system owns each field, so the two never overwrite each other.

Can the agent use data from our ERP, help desk or inbox?

Yes, and this is often the reason to build a custom agent. It can check open invoices in QuickBooks or NetSuite before a renewal call, read Zendesk tickets before a deal review, or attach a quote from your ERP to the opportunity. Each extra system gets its own permissions, read-only unless the job needs to write.

Rules & Data

What approval rules can we set?

You decide per action. Common rules: drafts always need a rep to send them. Stage changes need the deal owner's approval above a deal value you choose. Empty fields may be filled without approval, but a field a person typed is never overwritten. Merges and deletes are not allowed. Every action is logged with the old value, the new value and the reason.

Is our CRM data safe with an AI agent?

The agent gets its own user with the fewest permissions the job needs, so it cannot see or change objects outside its scope. API keys stay in your accounts. The AI model can run in your own Anthropic, OpenAI or cloud account, so records pass through a contract you control. HubSpot's own MCP server also leaves out properties marked as sensitive data.

Which AI model do you use, and do you build CRM agents on Claude?

We build on Claude, GPT or Gemini and pick per task, and many of our agents run on the Anthropic API. With tool use, the model can only ask to run functions we define. Our code checks each request against your rules before anything reaches the CRM. The model sits behind that code layer, so you can switch providers later without a rebuild. FactoryJet is an independent services company, not part of Anthropic.

How do we know the agent is doing a good job?

Every action is logged with the reason the agent gave. In the first weeks we review a sample with your sales or RevOps lead: how many suggestions were accepted, how many were edited and how many were wrong. Rules tighten or loosen based on that review, and the numbers come from your own CRM.

Process

How long does it take to build a CRM AI agent?

A single focused job, such as lead qualification in HubSpot, usually goes live one to three weeks after the scope is signed, the same range we give for our other AI integrations. An agent that covers several jobs or two CRMs takes longer. The written scope has dated milestones, and you see a demo inside your own CRM every week.

What do you need from us to start?

A short brief: which CRM and edition you use, the job you want done first, how many leads or deals move each week, and who approves the agent's actions. Test or sandbox access helps. If you already have an RFP or a requirements document, send it as it is. We answer the questions in it, so you do not have to fill in a form of ours.

Can we see it work before we sign?

Yes. Before you sign, we run the agent on a small set of your real records, with names removed if you prefer. You see each decision it would make and the reason for it, where it hands off to a person, and exactly which fields it would change in the CRM.

What happens when HubSpot or Salesforce changes its API?

Both vendors version their APIs and announce changes ahead of time. The agent calls the CRM through one small layer of code, so a change is usually fixed in one place. Failed calls send an alert to Slack or email. If you choose support, we watch those notices and quote any needed change before doing it.

Cost & Ownership

How much does a custom CRM AI agent cost?

Market figures vary widely. Development firm ProductCrafters, in the FAQ of its February 2026 cost guide, puts custom-built AI agents with integrations at $15,000 to $75,000, and enterprise multi-agent systems higher. FactoryJet does not publish a price list. We quote a fixed price per phase in writing after a short scoping call. Later changes are quoted and approved one request at a time.

Are there monthly fees after launch?

Not from us unless you choose support. The AI model and hosting can run in your own accounts, billed to you at cost with no markup. Defects found after launch are fixed inside the build price for a cover period agreed before work starts. If you choose support, we keep managing the servers, model bills, updates and security, quoted as a fixed price. Each change is quoted and approved first, and there is no monthly minimum.

Who owns the agent and its code?

You do, from day one. The code lives in a repository you own. The HubSpot app or Salesforce connected app is registered in your account, and the prompts and approval rules are written down in plain English. If you bring the work in-house later, your team has everything it needs to run and change the agent.

SOURCES FOR THE ANSWERS ABOVE, CHECKED 31 OCTOBER 2026

SEE IT ON YOUR RECORDS

Watch the agent work on your own CRM data before you sign

Send the first job you want handled. We run it on a sample of your HubSpot or Salesforce records and show you every decision and its reason.

See all AI services

Demo before you sign. Fixed price per phase. You own the code. No monthly minimum.

Free quote
Founder replies in 24h