Custom AI agents that handle enquiries, quotes and follow-ups
They work inside your email, calendar and CRM, so customers hear back quickly and quotes go out on time.
Updated October 2026Industrial AI builds custom AI agents for New Zealand businesses that reply to enquiries, prepare quotes, sort the inbox and follow up, working inside your own email, calendar and CRM. Each agent stays within the limits you set and passes anything unusual to a person.
- Works in the tools you already use
- Takes a task from start to finish
- You choose what needs your approval
- Hands over to a person with the full story
- Starts with one job, then grows
Follow a quote request from start to finish
Each agent has one clear job, access to the tools it needs, and rules you set.
The same approach works for bookings, supplier emails and maintenance requests. For phone calls, see the AI voice agent.
- 08:14 In
A quote request comes in
From the website form: “Could I get a price for a regular office clean, and when could you start?”
- 08:14 Read
Reads the details
Picks out the job, location and contact details, and adds the new customer to your CRM.
- 08:15 Check
Prepares the quote
Prices the job from your own price list and drafts the quote in your template.
- 08:21 Approve
Asks for your OK
Your rule: quotes for new customers need approval. You approve it from your phone.
- 08:22 Send
Sends the quote
Emails the quote with two times that suit your calendar, and logs it in the CRM.
- Day 3 Follow up
Follows up
No reply, so it sends a friendly follow-up. If there’s still nothing, it lets you know.
Work an agent can take on for you
Replies to enquiries
Answers web, email and message enquiries from your information, and qualifies the lead.
Prepares quotes
Priced from your list, sent in your template and followed up on time.
Sorts the inbox
Files routine mail, drafts replies and flags what needs you today.
Books appointments
Makes, moves and cancels bookings in your calendar.
Handles multi-step jobs
Reads a request, decides the next step and updates each system involved.
Knows when to hand over
Passes anything unusual to the right person, with the full history.
How an AI agent differs from a chatbot
Answers questions
- Answers a question, then stops
- Can’t see your calendar or systems
- Leaves the next step to you
- Forgets context between steps
Gets the task done
- Connected to your tools, with permission to act
- Books, quotes, updates and follows up
- Keeps track of the whole job
- Hands over to a person with the full history
Agentic AI for New Zealand business
Agentic AI works towards a goal. It plans the steps, uses your tools and checks the result, instead of only answering a question.
A chatbot replies and waits for the next message. An agentic system takes a request such as “quote this job and book a site visit”, works out the steps, looks up prices and availability in your systems, drafts the quote, books the visit and reports back. It uses the same language models as chat tools, given access to your tools and a defined job.
Where it works well
- Requests that vary but follow a familiar pattern, such as enquiries, quotes, bookings and supplier emails
- Work that moves information between systems, like an inbox, a CRM and a calendar
- Follow-ups that slip when the team is busy
Where it needs a person
- Decisions with legal, financial or safety consequences
- Situations it can’t check against your own information
- Anything where a wrong answer costs more than the time saved
Dependable agentic AI works within firm limits. Each agent gets only the tools its job needs, waits for approval at the points you choose and logs every action, so your team can see what it did and why.
You stay in control of what it does
We agree the rules with you before go-live, and you can change them at any time.
- Limited access: only the tools and records its job needs
- Approval points: you choose what it does alone and what it asks about first
- Action log: a record of what it did, when and why
- Hand-over: anything unusual goes to a person, with the history
- Privacy: your data kept separate from other businesses
Start with one job and prove it
Works with tools like these
- Gmail & Google Workspace
- Outlook & Microsoft 365
- HubSpot
- Pipedrive
- Xero
- MYOB
- ServiceM8
- Slack & Teams
- Spreadsheets & forms
- Website chat & SMS
- 01
Pick the first job
We choose one workflow worth handing over, the tools it touches and where you approve.
- 02
Build and connect
The agent is set up in your systems, with its limits agreed.
- 03
Test on real examples
It’s tested on real examples before it deals with live customers or records.
- 04
Hand over and grow
Your team learns to work with it, then we add the next job.
Best for businesses with a busy inbox
If enquiries, quotes and follow-ups arrive faster than your team can get to them, an agent can help.
- Trades and home services
- Property and real estate
- Clinics and professional services
- Retail and hospitality
- Manufacturers’ sales and admin
When to use an agent, a voice agent or automation
An AI agent suits requests that vary and need some judgement. Work that comes in by phone suits the AI voice agent, and fixed, repeatable processes suit business automation. For a private assistant you message on WhatsApp or Telegram, see OpenClaw setup or your own AI employee.
To plan first or train your team, see AI strategy, the AI agent workshop and AI workshops.
Frequently asked questions
What is an AI agent?
What is agentic AI?
How is an AI agent different from a chatbot?
What can an AI agent do for my business?
Will it act without asking us?
Is our data safe with an AI agent?
Which AI model does it use?
How is it priced?
More on AI agents
Talk to us about your enquiries
Tell us where work piles up, and we’ll suggest the first job for an agent.