The World Economic Forum’s report states that 34% of all business-related tasks are currently performed by machines, and the trend of automating operations continues to grow with the rise of AI agents. In a world where AI agents are revolutionizing the way we work and interact, the concept of AI agents is no longer a futuristic dream but a tangible reality.

In a recent episode of the Invisible Machine podcast, Annie Harshberger, Lead Experience Architect at OneReach.ai, demonstrates multiple AI agents swarming around a complex scheduling automation. Together with the bestselling authors of Age of Invisible Machines, Robb Wilson (CEO and Co-Founder of Onereach.ai), and Josh Tyson, they explore how these agents communicate with each other to get work done in real time—setting up a meeting with all the necessary details such as agenda, participants, and expected outcomes. 

Let’s delve into the details of using a multi-agentic approach for scheduling meetings, managing calendars, and more.

Episode highlights:

• Streamlining the scheduling process
• Coordination of multiple AI agents through a single entry point
• Constructing AI agents with reusability in mind
• Predicting the best options for users based on context and conversation history
• AI agents functioning as a swarm to complete tasks

Collaborative AI Agents: Compartmentalizing the Tasks

I think the way we were thinking about it is looking at those agents in terms of reusability and how certain pieces of that experience will be used not just when you’re scheduling a meeting but also in other skills.

— Annie Harshberger, Lead Experience Architect at OneReach.ai

In developing AI agents, or intelligent digital workers (IDWs), our product team stresses the importance of reusability, ensuring that components of the scheduling experience can be applied to various tasks beyond just setting up meetings. For instance, the PollTaker agent that collects voting results to determine optimal times for a group meeting can easily be used by another agent working on the objective of gauging the adoption rate of a new employee benefits program.

Our Rich Web Chat (RWC) product enables access to various agents through a single entry point, which drastically reduces task-switching. Authentication is the first crucial step, identifying the user to ensure the agent executes tasks in context, knowing who it’s performing tasks for. This context includes the user’s past interactions and meeting history, which are specific to the logged-in user.

Scheduling requires several agents working together behind the scenes:

• The SchedulerCoordinator agent collects essential meeting details like the subject, agenda, and expected outcomes.
• The MeetingDetailsCollector agent gathers and handles all necessary information from the user.
• The TimeSyncer agent proposes time slots and manages participants.
• The NotificationSender agent prepares and sends messages to participants.
• The Reminder agent sets up timely reminders for the meeting.
• The ParticipantsResponseChecker agent processes all voting data, identifying response patterns and predicting optimal times based on historical data.

The predictive capability of AI agents aims to enhance the user experience by suggesting optimal times even before soliciting votes. Such AI agents can be very useful in, for example, setting up a parent-teacher conference or filling last-minute cancellations at a doctor’s office.

Maximizing Efficiency with AI Agents

The power of AI agents lies in their ability to function both independently and as part of a swarm, to efficiently manage and execute complex tasks while staying close to human set objectives and the humans who set them. An IDW can actually manifest as a swarm of different agents handling several tasks simultaneously, such as sending out meeting invites to different participants. This ability to run in parallel amplifies productivity and ensures that tasks are completed swiftly and accurately. 

In the demo, we get to see each AI agent as it is called on, and while this level of detail isn’t critical to end users, it does show how many different points in the task would require moving to a different tab or app in a manual approach. 

By understanding and utilizing these dynamic AI agents, businesses can significantly streamline their operations, enhance efficiency, and achieve better outcomes. The future of work is here, and it’s driven by the intelligent collaboration of AI agents.

Watch the full episode for more insights on YouTube or listen to it on Apple Podcasts/ Spotify.

Do you want to learn more about orchestrating AI agents on the OneReach.ai platform? Download our whitepaper, “What Everyone Is Getting Wrong About AI Agents”.