Atlassian Intelligence suggest request types: A complete overview

Let's be honest, setting up a new service project can be a real drag. Before you can actually start helping people, you have to try and predict every single thing they might need. It's a long process of defining request types, building out forms, and crossing your fingers that you didn't forget something important. For teams just getting started with ITSM, it often feels like a lot of guesswork.

Atlassian is trying to take some of that pain away by building AI features straight into its tools. One of the most talked-about is the ability for AI to suggest request types in Jira Service Management (JSM).

This guide will walk you through everything you need to know about the feature. We’ll cover what it is, how it works, where it shines, and, maybe most importantly, the limitations you should be aware of. We'll also look at how a more specialized AI platform can pick up where Atlassian's tool leaves off and really level up your service desk.

What is Atlassian Intelligence Suggest Request Types?

First, a quick intro. Atlassian Intelligence is the name for a bunch of AI-powered features popping up across its cloud products like Confluence, Jira, and JSM. The idea is to give you a "virtual teammate" to help you get work done faster.

The Atlassian Intelligence Suggest Request Types feature is one of these tools, built specifically for Jira Service Management. It uses a simple text prompt from an admin to generate a list of request types for a service project. So, instead of brainstorming every possible request yourself, you just describe what your team does, and the AI gives you a starter list.

Its main job is to speed up the initial setup of your service catalog, making it easier to create a clear portal for your customers or employees. It’s part of a bigger AI push in JSM that also helps with things like suggesting form fields and sorting incoming tickets.

How the Suggest Request Types feature works

Atlassian's AI is designed to help you out at two main points: when you're building your service desk for the first time, and when you're managing the day-to-day ticket queue. Here’s a look at how it plays out.

Generating new request types

This is where the feature really comes in handy, especially when you're starting from scratch. If you're a JSM admin, the process is pretty simple:

1. Head over to Project settings > Request management > Request types.

2. Click Create request type and pick the Create using AI option.

3. A text box will pop up. This is where you describe your team. For example, you could type, "We are a facilities team that handles maintenance requests, office supply orders, and meeting room bookings."

4. Atlassian Intelligence takes your prompt, thinks for a second, and then spits out a list of suggested request types, like "Request building maintenance," "Order office supplies," or "Book a conference room."

5. You can then look over the list, pick the ones that make sense for your team, and create them in one go.

The suggestions usually come with a few basic system and custom fields already attached, giving you a template to work from.

Suggesting relevant fields for your forms

Once you have your request types, the next job is to build the forms to go with them. Atlassian Intelligence can lend a hand here, too. After you create a request type (whether with AI or on your own), you can use the Suggest fields feature.

Based on the name and description of the request type, the AI will recommend which existing fields to add or even suggest new custom fields to create. For a "New Laptop Request," it might suggest fields for "Department," "Required Software," and "Business Justification." This helps make sure your team gets all the info they need right away, cutting down on the back-and-forth emails.

Triaging and updating existing requests

The AI's help doesn't end once you're set up. It can also help with managing your ticket queue through AI-powered triage. This is especially useful for requests that come in through email, since they often just get dumped into a generic "Emailed request" bucket.

You can select a bunch of these un-sorted tickets, and Atlassian Intelligence will read through them to suggest better request types. For instance, an email with the subject "My mouse is broken" could be correctly moved to the "Report a hardware issue" request type. You can review and apply these changes in bulk, which saves your team a ton of manual sorting.

The limitations of Atlassian Intelligence Suggest Request Types

While Atlassian Intelligence gives you a decent starting point, its built-in nature comes with some real trade-offs that can hold you back. It’s a bit like a multitool, handy in a pinch, but you wouldn't use it to build a house.

It only works within the Atlassian ecosystem

The biggest issue is that the AI is stuck in a walled garden. It can only learn from your Atlassian products, like Confluence pages and Jira tickets. But think about it: where does your team actually keep its knowledge?

Most companies have important info spread all over the place, in Google Docs, on internal Notion wikis, in old conversations on Zendesk or Intercom, and buried in countless Slack threads. Atlassian Intelligence can't see any of that. This means your AI is working with one hand tied behind its back, leading to half-baked suggestions and an inability to answer questions that need info from outside Atlassian. You're left with two not-so-great options: either spend ages copying everything into Confluence, or just accept that your AI is missing most of the picture.

This infographic shows how eesel AI connects to all your knowledge sources, unlike the siloed approach of some built-in tools.

Lack of advanced customization and control

The simplicity of the prompt system is also its biggest weakness. You can't train the AI on your team's specific tone of voice from past tickets, and you can't set up complex, multi-step workflows.

The automation is pretty basic, limited to filling in fields and changing categories. You can't, for instance, create an action that lets the AI look up an order status in Shopify, check a user's subscription level in your own database, or escalate a ticket because it detected an angry customer. This keeps the AI in a purely advisory role, stopping it from taking the kind of helpful, independent actions that actually resolve issues and free up your team.

This image shows the eesel AI interface, highlighting the powerful customization and workflow rules you can create.

Basic simulation and reporting features

For teams that care about getting things right the first time, this one might be the most concerning. Atlassian's AI features are turned on directly in your live project. There isn't a proper sandbox or simulation mode where you can safely test how the AI will behave with your real data before you let it loose.

This "flip a switch and hope for the best" approach is risky. You have no real way to predict its accuracy, how it will affect your workflows, or if it might accidentally miscategorize a critical incident. You only find out what's wrong after it's already touching your tickets. For anyone who needs to be sure their automation is reliable from day one, that's a pretty big deal-breaker.

A screenshot of the eesel AI simulation mode, which allows you to test AI performance on historical data before deployment.

A more flexible alternative: eesel AI

For teams that want to get past these limitations, a dedicated AI platform like eesel AI offers a more powerful, connected, and safer way to do things. It plugs right into your tools, including a seamless Jira Service Management integration, to bring you automation that actually works the way you do.

Unify all your knowledge, not just some of it

Unlike Atlassian's walled garden, eesel AI is built to break down information silos. It connects to all your apps right out of the box, learning from Jira tickets, Confluence, Google Docs, Notion, and even past helpdesk conversations from places like Zendesk.

This means your AI agent is trained on a complete picture of your company's knowledge. The result is much more accurate and context-aware automation, whether it's answering questions or triaging tickets. eesel AI can even analyze your best ticket resolutions and automatically draft new knowledge base articles, helping you fill information gaps with content that you already know works.

Gain total control with a customizable workflow engine

With eesel AI, you're in full control. A powerful prompt editor and workflow builder let you define the AI's exact personality, tone of voice, and the specific actions it's allowed to take.

Custom Actions are where things get really interesting. They let your AI agent connect to any external system through an API. It can look up order details in Shopify, verify a user's permissions in your internal admin panel, or create a new issue in a separate developer project. This turns your AI from a simple assistant into an agent that can solve problems from start to finish. You get to decide exactly which ticket types the AI handles, so you can start small and scale up your automation as you get more comfortable.

Test with confidence using powerful simulations

One of the biggest advantages of eesel AI is its simulation mode. Before you turn anything on, you can run your AI setup against thousands of your past tickets in a safe, sandboxed environment.

This shows you exactly how the AI would have responded to real requests from real people. You can get accurate predictions on resolution rates, spot gaps in your knowledge base, and tweak its behavior before it ever interacts with a customer. This risk-free testing lets you deploy with confidence and gives you clear reports on what to improve next.

Atlassian Intelligence Suggest Request Types pricing

So, what does all this cost? Atlassian Intelligence features, including Suggest Request Types, aren't part of every plan. They are only bundled with the Jira Service Management Premium and Enterprise cloud plans.

This means if you're on the Free or Standard plan, you won't be able to use these AI tools without a pretty big subscription upgrade. The pricing is per agent, so the cost can add up fast as your team grows.

Heads up: Prices are approximate and can change. Check the official Atlassian pricing page for the latest info.

Atlassian Intelligence Suggest Request Types: A good start, but not a complete solution

Atlassian Intelligence Suggest Request Types is a genuinely helpful feature for teams already deep in the Atlassian world who need a quick way to get a JSM project going. It makes getting started a little less intimidating and can definitely save you some time on the initial setup.

However, you can't ignore its shortcomings. Because it only works within its own ecosystem and lacks deep customization, solid testing features, and true workflow automation, it feels like an incomplete solution. For teams serious about using AI to deliver fast, smart, and reliable service, a built-in tool probably isn't going to be enough.

For a truly powerful, flexible, and safe AI layer that works with all of your tools, a dedicated platform is the way to go. See how eesel AI can transform your Jira Service Management workflows today.