According to IDC statistics, employees typically spend 2.5 hours a day looking for the right data. To find general information, we often use search engines such as Google or Bing. However, the search experience and the ability to quickly find relevant information within the internal organization are often much worse. Reason for this municipality to experiment with a renewed search solution that intuitively supports the employee in finding the right information.
One of the first things people often do when they're looking for information is to ask colleagues about the information, or where to find it. Through the interaction that takes place, the search question becomes increasingly clear and it becomes clear what exactly people are looking for. But what if you can ask questions to a chatbot that helps you in your search? Consider, for example, the conversation you can have with LLMs such as ChatGPT when you're looking for specific information. Equalling the most intuitive form of user experience, namely interaction through a chat robot, offers opportunities.
However, using LLMs in the search for the right answer is not without risks; something LLMs often have trouble with is telling where information comes from or referring to the right source. Querying the system insinuates that there is an answer, and that's why LLMs like ChatGPT won't hesitate to come up with a source or facilitate you from an answer from a document whose title is “wrong answers.”
Municipality officials can now use an internal search chatbot that assists them in finding the right information to do their job properly. By interacting, the search chatbot interprets exactly what information the employee is looking for, making a selection of links, separate documents with information, or domain experts. Similar to the search functionality as you know it from search engines. The employees describe their search, after which the customized AI system presents a number of resulting options. Is the answer not listed? Then, instead of trying a changed search, the user can start the conversation and receive new options.
In addition, the search chatbot can formulate an answer with information from various sources. In short, officials can use the search functionality through a conversation, also known as “conversational search”.
This renewed way of searching shifts the emphasis from searching for information to applying knowledge to the execution of work processes. For example, finding information is no longer a time-consuming task, but knowledge is just one conversation away from the search chatbot.
.png)