Updated on Jun 13th, 2024

Need to Know 

  • Many Public Service clients need sophisticated conversational agents rather than basic chatbots to save time and costs, and improve user experiences.
  • Conversational agents use generative AI for advanced interactions, handling tasks such as search functions, automated services, and complex data parsing, far surpassing the capabilities of traditional rule-based chatbots.
  • When built thoughtfully, conversational agents can save organizations significant time and money. One of our municipal clients saved thousands by implementing a website and phone agent that solved more than 70% of all user queries.

In recent years, one of the most frequent requests we've received from our clients has been to install chatbots on their websites. However, our Public Service clients in governments, cities, and public agencies often end up requiring a more sophisticated conversational agent.

Keep reading to learn about the exciting use cases for AI-enabled conversational agents, and how they are helping Public Service organizations save time and costs while supporting better experiences for users and internal teams. 

What’s a chatbot vs. a conversational agent?

A chatbot typically refers to a rule-based application programmed to reply to user inputs based on predetermined conversation flows. Although chatbot technology has evolved quite a bit over the decades, even the most advanced technology often still can’t hold up in conversation with real-life humans.

Conversational agents on the other hand use language models powered by generative AI. They’ve evolved past rule-based applications, allowing for more sophisticated conversations and applications, including search functions, automated services, and parsing complex data.

You've likely seen a chatbot icon hovering in the bottom right corner of a website many times before, but agents are no longer restricted to a pop-up modal window.

We have a municipal customer with a contact centre that was constantly overwhelmed. Their calculations indicated that each call cost over $11. Many of those calls involved information that was readily available on the website, but residents were choosing to call their office instead of searching for the content online.

We identified two key areas generating nearly 40% of daily calls and delivered two conversational agents for their website and 311 system.

These agents solved more than 70% of the queries residents had, saving the municipality hundreds of thousands of dollars in the first three months alone. 

What are the use cases for a conversational agent in Public Service organizations?

Some additional use cases for conversational agents in Public Service include:

  • Bus Stop Agents: Riders can text their stop ID, and the agent will respond with the estimated arrival time of the next bus. The agent can also answer questions about fares, trip planning, and bus routes.
  • Recreation Agents: Facility management software for cities and townships can be complex. A conversational agent can be trained with information on all facilities and schedules to answer resident questions, including broader inquiries, like for example, “Where are the open swim lanes right now?”
  • Open Data Agents: Training an agent on an open data catalogue and connecting it to a large language model with generative AI capabilities can allow residents to ask plain-language questions and get accurate, plain-language answers.
  • IT Support Agents: Larger organizations usually have an internal IT support desk, but as many of us have learned the hard way, IT issues don’t always happen between 9 a.m. and 5 p.m. Using support desk call logs, internal wikis, and software product documentation, many organizations have built self-serve conversational agents to support their teams 24/7.
  • Job Board Agents: Many government agencies use third-party job posting tools that can be difficult and time-consuming to navigate. We’ve seen great results for our clients that have trained conversational agents using their own job postings and HR policies to connect potential employees with the roles that are best suited to them.
  • Appointment Booking Agents: Aligning schedules presents a challenge in various sectors, including healthcare, permits and licensing, inspections, and other appointment-based industries. Training a conversational agent on an organization’s subject matter and integrating it with calendars enables users to call or text the agent to book a timeslot. The agent will then add the meeting to the calendar and can send an invitation to the user.
  • Multilingual Agents: One of the benefits of conversational agent platforms is that they are often multilingual. Building an agent that is trained on a specific data set and enabling multilingual capabilities will allow newcomers or people who don’t speak an organization’s primary language to ask questions in the language they are most comfortable using.
  • Event Agents: Finding out what’s happening in a specific community can often be more challenging than necessary. Training an agent on multiple event calendars and data sources can help organizations create a “what’s happening” search built around conversational AI.
  • Ask an Expert Agent: Organizations that contain many areas of specialization can create agents trained on not just their web content, but the many PDFs, policies, and even “dark web” content in an internal search not available to the public can help employees find answers fast. 

Where should organizations start if they want to build their own conversational agent?

Here are some approaches to start formulating how a conversational agent could benefit your organization:

  • Speak with your team about the questions they are asked most frequently. If you organization has a call centre, or a collection of call logs, these can be good places to start. Finding questions that a conversational agent can handle for your team can help you reduce hold times and save costs.
  • Research your internal site search logs. Go to your web analytics and look at what people are searching for on your site. This can give you an understanding of where users are getting lost or the journey is unclear on your site. If there are topics that users aren’t finding but that you have strong content or data for, an agent can be used in this instance to help bridge the gap.
  • Check your web analytics for inbound search terms from external search engines. Review the user journeys for the top terms to see if they are being connected with the information they are searching for. In situations where they aren’t, these are the journeys you can mine for opportunities.
  • Talk to your front-line employees who interact with the public. They will have anecdotes and stories about what customers need on a daily basis, opening up additional avenues to explore.

As you and your team identify challenges, the following three questions can help you determine if a conversational agent is the right solution:

  1. Do we believe a conversational agent can save end users time or steps in this situation?
  2. Do we have the right content or data to train the agent? Is it of good quality?
  3. Are these problems straightforward? It’s fine to use an agent for complex tasks later on, but starting simple to allow your teams to get up to speed is best.

Thinking about implementing a conversational agent for your organization? Our experts would love to help you explore how the right build can make a measurable impact on your user experiences. Contact us today to get started.