Mastering Jira Virtual Agents: Flows, Intents & More

A quick introduction into what a JSM virtual agent is and what we can expect it to do.

Intents are a customers goal - what they want to do. Creating good intents, and ensuring they're backed up by good training phrases, is crucial to having a well-working agent.

Virtual agents use flows to guide their conversation. Flows tell the agent to send a specific message, solicit information, or take some action.

Every virtual agent and evert intent shares these standard flows. These include things like a greeting, intent-matching and escalation. While we can't edit most of them, we can tailor them to our needs.

We'll take a look at other settings and features available to JSM virtual agents, including things like:

1. Atlassian Intelligence answers - let your agent get answers from Confluence

2. Changing the default request type - The agent must have a default request type, but you can change it

3. Connecting other channels - Let customers interact with your agent via slack, teams and more

Having a virtual agent is a great first step, but monitoring it for success is even better. Here we'll dig into what metrics are available in Jira Service Management virtual agents.

I'll quickly go over what we'll be doing in this hands-on section.

Be sure to read the next lecture (Our Scenario) to get important information if you're following along!

Note to follow along in your instance you must :

1. Be on Jira Premium or Enterprise

2. Be a project admin

We'll use this scenario to guide our virtual agent in this hands-on section. Feel free to modify, or even use your own as you follow along!

We'll get hands-on in Jira Premium and create a virtual agent.

We'll examine the standard flows and see how we can tailor them to better meet our teams needs.

Remember to refer to our scenario at the start of this section!

We'll take the agent we've created and think up some intents and training phrases we might want to use. Remember to refer to our scenario at the start of this section!

Here we'll become a customer and interact with the agent to see what it looks like.

We'll wrap up this section by looking at the metrics created by our agent. We'll also brainstorm what changes we could make to the agent to make it more effective.