Customer Support and Voice AI
AI receptionist for small business: calls, bookings and handoffs
What an AI receptionist does on a real call, the rules that apply, a sample greeting and a pass/fail checklist to launch one safely.
Support Operations
AI Support Copilot for Zendesk
Zendesk agents open a ticket and find a drafted reply, the right macro and the source article already waiting for them.
Zendesk agents spend precious minutes per ticket hunting for the right macro and article. This copilot drafts a reply for each ticket an agent opens, grounded in your help center articles, macros and policies. It names the macro that fits, links the source article, and leaves everything as an internal note. The agent edits and sends; the copilot never replies on its own.
- A grounded draft on every Zendesk ticket an agent opens
- The right macro named and the source article linked
- Drafts stay internal notes until an agent sends them
$5,900
20–30 working days
Optional maintenance $295/mo
What it is
AI Support Copilot for Zendesk drafts a suggested reply for each ticket an agent opens, grounded in your help center articles, macros and policies. It names the macro that fits, links the source article, and leaves the draft as an internal note for the agent to edit and send. It never replies itself.
Why businesses buy it
In a busy Zendesk account, handle time is driven by searching, not typing: which macro, which article, which policy applies. That search repeats on every ticket and varies by agent. After launch, each ticket arrives with a draft, a macro suggestion and sources, so agents reply faster and more consistently. The Scope Card lists the Zendesk fields, sources and approval rules before you order, so you know exactly what the copilot writes and that customers only ever hear from your people.
Problems this solves
Zendesk agents start every reply from a blank box
For each ticket an agent opens, the copilot drafts a suggested reply grounded in your help center articles, macros and policies. The draft appears as an internal note.
The right macro exists but agents do not remember it
The copilot names the macro that fits the ticket and links the source article. Agents apply what your team already approved instead of rewriting it.
Support leads want AI help without automatic replies
The copilot never replies itself. Each draft stays an internal note in Zendesk until an agent reviews, edits and sends it, inside your approval rules.
A good fit if
Zendesk agents search the help center and scroll through macros on every ticket before they write a word.
Your macro list has grown long, and newer agents rarely know which one fits the ticket in front of them.
Two agents answer the same question differently, and you want every draft built from the same approved articles.
Your admin can create a Zendesk API token with an agent or admin role and give thirty minutes to a setup call.
Not the right fit if
You want routine tickets resolved without an agent involved. See AI Customer Support Agent for Zendesk.
Tickets reach the wrong group before anyone drafts anything. See AI Ticket Routing for Zendesk.
Your help center has gaps the copilot would keep running into. See AI Support Knowledge Agent for Zendesk.
You want finished conversations scored against a quality rubric. See AI Support QA.
How it works
- On a thirty-minute call your Zendesk admin and we agree what the copilot reads and writes: tickets, user records, macros, help center articles, and where drafts appear.
- When an agent opens a ticket, the copilot reads the thread and the requester's user record for context.
- It drafts a reply grounded only in your help center articles, policies and approved documents, up to 100 in total.
- Where one of your macros fits, it names that macro, so the agent can apply standard wording instead of retyping it.
- The draft and links to its source articles are added as an internal note. The agent edits, sends or ignores it, and nothing reaches the customer automatically.
- Actions such as a refund or account change are prepared for the agent to confirm under your approval rules. Fifty test scenarios on test tickets run before launch.
What you provide within 72 hours
Your portal lists these the moment you order. Delivery starts the working day after everything is in and your access is checked.
- A Zendesk API token with an agent or admin role, created by your admin.
- Your Zendesk admin for the thirty-minute mapping call, and test tickets if you have them.
- Your help center articles, FAQs and policies, plus the macros agents use most.
- Your playbook and approval limits for refunds or account changes, and the approver for exceptions.
- Ten real tickets with the replies your best agents sent.
- Your escalation rules and support hours.
- One decision-maker who approves draft style and scope.
- An API key for your own AI model account (for example OpenAI or Google Gemini), created for this build with a spending limit.
The 72-hour rule. Send everything on your checklist, including access, within 72 hours of ordering so your build starts on time.
Timeline
Start
Intake checked; Zendesk admin call booked.
Weeks 1–2
Mapping agreed; drafting built on your articles, macros and policies.
Week 3
Zendesk and second system connected; internal-note drafts tested.
Weeks 3–4
Fifty ticket scenarios run, with drafts shared for review.
Week 4
Agent feedback; two rounds of changes.
Go-live
Live beside your agents; recorded 60-minute handoff; support starts.
| Option | Delivery time | Price |
|---|---|---|
| Standard | 20–30 working days | $5,900 |
| Express | 15–25 working days | $6,800 |
| Rush | 10–20 working days | $7,650 |
What you receive at handover
- The support copilot working inside your Zendesk account, posting drafts as internal notes
- Two-way Zendesk connection with documented field mapping, plus up to one more system
- Up to six workflows with your approval rules, using your help center, macros and policies
- Access through invites or keys you control, with no keys sent by email
- Test report covering 50 scenarios on sandbox or test records
- Admin guide, recorded 60-minute handoff session and 30 days of post-launch support
Expected benefits and how to measure them
- Agents can reply faster, starting from a sourced draft.
- Helps put your existing macros and articles to fuller use.
- Every customer reply still passes through a person.
- Full resolution time
- Use Zendesk Explore to compare full resolution time for four weeks before and after launch.
- Macro usage
- Compare macro application counts in Zendesk Explore for the month before and after launch.
- Drafts used by agents
- Sample tickets each week to see how often the internal-note draft formed the basis of the sent reply.
Before you start: Save Zendesk Explore figures for resolution time, macro usage and tickets per agent for the month before setup.
These are things to track, not promised results. Outcomes depend on your data, your team and your market.
Running costs
The build is paid once. Running it costs roughly $50 to $250 a month for AI model usage and hosting in your own accounts, on top of your Zendesk plan. More agents opening more tickets moves you toward the top of that range.
Usage during testing is billed to your accounts as well.
Ask about running costsCustom requirements
Drafts across several Zendesk brands, a third connected system, or more than six workflow types need a written Change Quote. Add a checkout note and we send it to you.
Request a custom versionKeep it running
Monthly maintenance covers monitoring, fixes and change hours for work such as new macros, updated policies or a new ticket type. It starts after the post-launch support period, bills monthly and can be cancelled any month.
$295/mo (5% of the build price a month, includes 2 change hours)
Optional. Starts after the post-launch support period. Cancel any month.
How maintenance worksHow this compares
| Do it yourself with a SaaS tool | Hire a custom agency | Upmark Labs | |
|---|---|---|---|
| Price | Built-in AI assist bills per agent seat | A Zendesk partner quotes per project | Ours is one fixed price, shown on this page before you buy |
| Macros | Agents search the list themselves | Depends on the brief | Ours names the macro that fits the ticket |
| Sources | May draw on anything indexed | Agreed during the project | Ours drafts only from articles and policies you approve |
| Delivery | Switches on quickly, then needs tuning | Set per project | Ours has a fixed delivery time shown before you buy, with faster options |
Questions about this product
Where does AI Support Copilot for Zendesk show its draft?
As an internal note on the ticket, unless your mapping agrees a different field. Agents see the draft, the macro it suggests and links to the source articles in the ticket they already have open. Internal notes are never visible to the customer, so nothing leaks before an agent decides.
Can the Zendesk copilot recommend which macro to use?
Yes. It reads your macros as part of the two-way sync and names the one that best fits the ticket, alongside its own draft. The agent applies the macro as normal. If no macro fits, it says so, which is also a useful signal about where your macro list has gaps.
Will the copilot ever post a public reply in Zendesk?
No. It only writes internal notes and, where your approval rules allow, prepares actions for an agent to confirm. Every public reply is sent by a person. That is the difference between this copilot and AI Customer Support Agent for Zendesk, which can resolve tickets itself.
What happens when our help center has no article for the question?
The copilot says there is no approved source rather than inventing an answer, and leaves the agent to write the reply. Those tickets are logged, so you can see which topics need articles. If gaps are common, AI Support Knowledge Agent for Zendesk drafts new articles from resolved tickets.
How much is the Zendesk copilot with express or rush delivery?
One fixed price covers the whole build and is shown on this page before you buy, with express and rush delivery listed beside it at a set extra cost. All three include the documented mapping, fifty ticket scenarios, two rounds of changes and a recorded 60-minute handoff.
Which second system pairs best with the Zendesk copilot?
Usually the system that holds what agents look up most, such as your CRM or an order system. With it connected, drafts can include account or order details rather than asking the agent to fetch them. The choice is made at intake, and a third system needs a Change Quote.
Guides on this topic
All guides Give Zendesk agents a first draft on every ticket.
Order the Zendesk edition and every ticket an agent opens carries a drafted reply, a suggested macro and source links, ready to edit and send.
Works well with
Support Operations
AI Ticket Routing
Routes tickets to the right team or agent based on content and skills.
Support Operations
AI Support Knowledge Agent
Finds gaps in your help center and drafts new articles from resolved tickets.
Support Operations
AI Support QA
Scores conversations against your quality rubric and highlights coaching points.