The Linux Foundation Projects
Skip to main content

Some use cases for interoperable conversational assistants

Our use cases revolve around situations when multiple assistants need to work together, either simultaneously or sequentially.

Some examples

Smart Errands

In this example, Emmett, a human, seeks assistance from Cassandra, his general AI assistant, to manage and streamline his errands efficiently. The AI assistants at various service points—Flora at Green Haven, Carl at the Post Office, ToolTime at the hardware store, and the Host at Thai Delight—facilitate the transactions. Emmett, a human, is running errands and he has the following goals:
• Find out if the local nursery has Norway Pines. He wants 8 trees.
• Mail a package at the post office.
• Pick up his chainsaw if the repair is finished at the hardware store.
• Order then pick up a carryout of Pad Thai from the local restaurant.
• Make his drive as short as possible.

Cassandra seamlessly identifies the appropriate assistants and directs the conversation to each one. The assistants transfer information between each other through the use of OVON messages.

Large Organizations

A large university has assistants that provide general information about the university such as the academic calendar, financial aid, parking, and history of the university. It also has dozens of academic departments, many with their own assistants, that provide information about the departments.

A student considering applying to the university might want to consult with the different assistants to find out both general information as well as information about specific academic majors. Currently, the prospective student can go to the university’s website and search individually for information about each topic. But then the student would both have to find the specific assistants as well as repeat a lot of information, like their high school GPA, whether they are in-state or out of state, or the degree they’re interested in. However, with interoperable assistants, the student would start with a general assistant, which could then direct the student to other assistants as needed. With interoperable assistants, the student doesn’t have to repeat information, because the information could be transferred throughout the conversation.

Even better, if the student wants information about something that the university assistants don’t know, for example, information about the local town where the university is, the university assistants could pass the student on to another assistant outside of the university system.

Government Services

Governments provide many services to their citizens, which are increasingly mediated through conversational assistants. For example, government assistants will soon enable citizens to access different government services, such as fishing licenses, food recalls, national parks, public libraries, and weather reports. Interoperable assistants could direct citizens to the assistants that can address their goals, no matter what government department is responsible for that information, or even if the assistant belongs to a non-government organization. This use case has actually been implemented by the Open Voice Interoperability Initiative in collaboration with the Government of Estonia. You can see what we did on this YouTube video.

Use cases for interoperable conversational assistants range from personal assistants that assist users in accomplishing everyday tasks by coordinating with assistants from the user’s services like local businesses, local government services or entertainment providers to large ecosystems of assistants provided by organizations like businesses and universities.