Welcome to the beta program for AI Builder Studio! 🥳

This document will guide you through the capabilities available in the current version of the AI Builder Studio and help you identify suitable use cases for your organization. We will continually expand our capabilities, drawing inspiration from your valuable feedback. Stay tuned for exciting updates!

What Can You Build with AI Builder Studio

The AI Builder Studio enables IT administrators to create AI Agents that simplify daily tasks and make them immediately available for execution via the Agent or End User Chatbot. During the beta phase, the Builder supports building automation blocks based on the following capabilities:

1. Azure Active Directory (Azure AD) Integration

Capabilities:

• User Management: Create, update, retrieve, or delete users in Azure AD.

• Group Operations: Manage group memberships, including adding or removing users.

• Authentication Control: Configure and manage password settings and access policies.

• Role Assignments: Assign directory roles to users for access control.

• Property Queries: Fetch specific user details, such as contact information or account status.

Example use cases:

• Automating user onboarding by creating users in Azure AD and assigning roles.

• Generating a list of locked/inactive users in Azure AD for cleanup.

• Managing group memberships for project-specific teams.

2. SysAid Analytics API

Capabilities:

• Data Queries: Run analytics queries with filters, groupings, and aggregations (e.g., count service requests by status or calculate resolution times).

• Predefined Lists: Fetch lists of statuses, priorities, or other predefined data.

• Data Models: Access information about service requests, user records, and assets.

  • Asset details include hardware specs, network information, location, ownership, maintenance history, and more.
    

Example use cases:

• Creating reports on service request resolution times.

• Automating inventory tracking for IT assets.

3. Large Language Model (LLM)

With LLM capabilities, you can build automation blocks that go beyond simple actions. These blocks can understand and work with natural language, making them ideal for analyzing and generating unstructured text.
This means you can build blocks that perform smarter tasks, such as understanding patterns in tickets, generating summaries, or automating complex analyses.

Example use cases:

• Duplicate Ticket Detection - Build a block for reviewing the content of the last 25 reported tickets to identify and flag potential duplicates.

• Automated Report Summaries - Build a block that reads a list of resolved tickets and generates a concise summary for a weekly report.

• License Request Processor - Create a block that collects relevant data for every SaaS license request (like costs, business justification, etc) before opening a ticket.

• Text Data Extraction - build a block that extracts key information (like error codes or affected systems) from ticket descriptions.

What You Can’t Build in the AI Builder Studio (for now 😉)

While the AI Builder Studio is already packed with powerful capabilities, there are still a few things it can’t do at this stage. Here’s what’s currently out of scope:

1. Scheduled or Triggered Processes

   Blocks created in the Builder can only be executed manually, either through a command in the Builder Studio or the agent chatbot. This means you can’t yet set a block to run on a schedule (e.g., every 2 hours) or trigger it automatically (e.g. when a new ticket is created).

2. Automation That Changes or Interacts with the User Interface

   The AI Builder Studio focuses on backend processes and doesn’t yet support automations that modify or interact directly with a system’s UI.

3. Data Retrieval from External Services (Beyond Azure AD)

   Currently, the only external service integrated with AI Builder Studio is the Azure Active Directory. It cannot yet query or retrieve data from other external platforms.

4. Building Full Integrations

   Although the AI Builder can generate code, it is not yet capable of setting up full integrations with external systems on its own.

How to Use the Blocks

Blocks created in the AI Studio Builder can be executed through the Agent or End User chatbot. Once you create and publish a block, it’s automatically added to a default dataset called “AI Blocks (AI Admin)” in the Agent chatbot’s Data Pool. From there, the block is available for execution via the Agent Chatbot as long as it remains published.

Planning Out Your Block

1. Identify Your Needs: Review the use cases provided above and consider areas where automation can save time or improve efficiency.

2. Experiment and Build: Use the intuitive chat interface to conversationally create action blocks and analytic blocks tailored to your workflows and operations.

3. Execute via Agent chatbot: After successfully creating your first blocks, publish them to be available for conversational execution through the agent chatbot.

4. Ask for Help: If you encounter challenges, reach out to the beta support team for assistance.

Building the block

If you want the block to be accessible from the End User chatbot, follow these steps:

1. Go to the End User chatbot’s Data Pool.

2. Create a new dataset of the type “AI Blocks.”

3. Use the Builder chatbot to add the block to this dataset 

Like with any other dataset, you can modify permissions on the blocks dataset to restrict or grant access based on user groups or companies.

Feedback and Iteration

Your feedback is crucial! As a beta user, you play a pivotal role in shaping AI Builder Studio. Please share your thoughts on:

• Ease of use

• Additional integrations or features you’d like to see.

• Any bugs or issues you encounter.

By starting with these foundational capabilities, you can explore impactful automation use cases and prepare for the advanced features planned in future updates.
If you have further questions or need support, feel free to contact us.

We're excited to see what you build with AI Builder Studio✨