What is a virtual agent?

Service desks almost always have a low level of “automation” that makes your life easier as an IT manager or IT professional. One such automation is the classic email response, “We have received your request and will get back to you shortly.” Enter the virtual agent. An AI-powered virtual agent is more than just a chatbot, thanks to technologies such use machine learning and natural language processing (NLP). This enables it to be an active participant in a conversation, acting more like a human. Your virtual agent can automate tier-1 support interactions, deflect common issues, and give you time back so you can focus on higher-value work. Beyond tier-1 deflection, today's virtual agents also support human agents directly - summarizing requests, triaging by severity, and surfacing the context needed to resolve complex issues faster.

Virtual agent vs. virtual assistant

Virtual agents are software that follow certain rules to provide answers or directions based on customer questions. How intuitive a conversation can be depends on the complexity of the ruleset. Jira’s new virtual agent enhances your rule set with AI, machine learning, and natural language processing. A virtual assistant is a person who works remotely, usually on administrative and transactional tasks. Virtual assistants help schedule meetings, make travel arrangements, and take and transfer phone calls. 

Benefits of a virtual agent

Virtual agents give time back to your team to focus on higher value work. A common task eliminated by using a virtual agent is triaging. IT professionals waste time tagging teammates and reassigning tickets before work can start. A virtual agent can get the right ticket to the right person at the right time. Beyond automated messaging capabilities, machine learning enables a virtual agent to communicate with your customers in a more human-like way. As soon as a ticket is created, a virtual agent begins working with your customers to identify the root of the issue. And when an issue needs escalation, the triaging and communication with teammates begin immediately.

Natural language processing

Natural language processing (NLP) is the ability of a computer program to understand human language, whether spoken or written. Your customers will give clues to how they are feeling and the severity of the issue with the words they use in a service request. NLP enables a virtual agent to determine severity and sentiment with greater accuracy. A virtual agent will use sentiment analysis to escalate issues and seamlessly transfer them to humans who are available to help. 

Voice assistants

Do your customers have a hotline they can call? Voice assistants can offer high levels of service by learning as much as possible about the customer and their issue before escalating to a teammate. Virtual agents elevate this experience by listening to how your customers respond and interpreting the tone of the conversation. Imagine a voice assistant that communicates the customer's issue and level of distress before you even pick up the phone.

How to use a virtual agent

Virtual agents follow rules to provide answers and manage communication based on customer inputs. Many service management teams already have flow charts and customer escalation processes that you can train your virtual agent on. Training and testing are a collaboration between the virtual agent and the team. The team needs to decide how they want to collaborate and ensure the virtual agent interacts with customers in ways that meet company standards.

Training a virtual agent

Training has long been the domain of data scientists and machine learning experts. Today’s virtual agents can be trained without code or specialized knowledge. Training involves digesting and analyzing a rule set and creating a model for future use. Once trained, teams should interact with the virtual agent in a sandbox environment. If you're unsatisfied with your test interactions, you can adjust and reconfigure your training data until your interactions are satisfactory. 

Integration with third-party channels

One of the challenges of service management is the many channels your customers use to reach you. Most service management software centralizes service requests from phone, email, text, Slack, and web channels. Teams then work on the requests and use the same service management software to communicate back to internal and external customers. What’s been missing is the software’s ability to match the tone, length, and conversational style of each third-party channel. A virtual agent solves this problem by deep integration with third-party communication channels and an intelligent understanding of how to communicate effectively in each channel. 

Conversational analytics

Conversational analytics is the study of social interactions that investigates what it takes to create mutual understanding between two parties. This branch of study has identified trends in the content and tone of conversations that help use better understand how satisfied our customers are. Rather than sending someone a customer satisfaction survey, conversational analytics gives you a real time look at customer satisfaction and identifies the moments where their satisfaction increases or decreases. Conversational analytics can serve as an escalation tool, notifying human teammates when the virtual agent needs help. 

Multilingual support

Google Translate is great in a pinch, but it is not sufficient for customer communication. Until recently, automated translation services mostly replaced words with their translations. This led to embarrassing contextual errors and poor communication. Using AI and NLP, virtual agents can assess the full intent of a statement and find the linguistic equivalent. A virtual agent can navigate natural language with more proficiency and allow you to serve more customers in more areas. 

Examples of virtual agents

Knowing what you know now, have you worked with a virtual agent? It’s possible that you’ve emailed a virtual agent without even knowing it. Understanding where and when virtual agents are being used successfully can help you decide the best uses for virtual agents in your business. 

Virtual agent for customer service

The goal of a customer service agent is to understand a customer's problem, determine the severity, and gauge their emotional state. A well-trained virtual agent can quickly interpret all three. Imagine a virtual agent running a software company's service desk. An outage begins, and customers immediately flood the service desk with tickets. Human customer service agents need to open each ticket before confirming that they are all related. A virtual agent can do that in an instant, fast-tracking the escalation of the issue. 

Virtual agent for sales and marketing

Cold emails and lead generation are time-consuming activities done by some of our most expensive employees. Imagine a world where a virtual agent actively learns about potential customers and engages them with personalized messaging. When an opportunity arises during the virtual agent's communication, a human salesperson is brought in to close the deal. 

Virtual agent for finance

Financial professionals spend an enormous amount of time compiling documents each quarter. These documents need to be structured the same way and communicate the same information. A virtual agent can do the heavy lifting for your finance professionals, freeing them to focus on the human side of financial reporting. Financial professionals are responsible for communicating results and projections to the business. A virtual agent can increase the speed and accuracy of these communications. 

Virtual agent for HR

HR professionals are excellent at helping employees reach their full potential. They are also tasked with annual and quarterly reporting on employee productivity and performance. If this process is software-based, it's an excellent candidate for management by virtual agents. A virtual agent that communicates with managers and completes reports helps your HR professionals focus on what they do best.

Virtual agents in Jira Service Management

Jira Service Management includes a virtual agent powered by Rovo, Atlassian's AI platform. The virtual agent operates inside Slack, Microsoft Teams, and the JSM customer portal, deflecting repetitive tier-1 requests - password resets, access provisioning, status updates, policy lookups - without requiring code. It draws on your existing knowledge base in Confluence, uses natural language processing to interpret each request in context, and either resolves the issue directly or hands it off to a human agent with the full conversation history attached.

Because Jira Service Management is part of Atlassian's Service Collection, a Rovo-powered virtual agent isn't limited to internal IT or HR use cases. The same bundle includes Customer Service Management for teams supporting external customers, letting you run a single AI service layer across IT help desks, HR service centers, and customer-facing support operations - all under one license.