Skip to main content

Interactive Demo

This interactive demo walks you through building and deploying an agent in Beak AI Studio, integrating data sources and skills.

Video Guide: Watch a step-by-step guide on YouTube: How to Create an Agent in Beak AI Studio

Scenario: Create and Deploy a Data Extraction Agent​

Step 1: Add Data Sources​

  • Navigate to the "Data Sources" tab.
  • Click "Add Data Source" and select a database, file, system or website that your agent will use.
  • Configure connection details and knowledge base settings for the datasource as needed.

Step 2: Add Skills​

  • Go to the "Skills" tab.
  • Click "Add Skill" to create or configure any pre-built or custom skills your agent will use (e.g., summarization, transformation).
  • Set skill parameters and input/output mappings as needed.

Step 3: Add Agent​

  • Go to Agents menu and click "Add Agent" under "My Agents" tab.
  • Enter an agent name, description and industry.
  • Select an agent architecture: choose from ReAct, Plan & Execute, Reflexion, Dynamic (adaptive), Multi-Agent Orchestration, or Multi-Agent Swarm. See Supported Architectures for details and guidance on selecting the best fit for your use case.
  • Provide appropiate instructions for the agent as needed.

Step 4: Test the Agent​

  • Use the integrated playground to run test inputs.
  • Inspect outputs, logs, and memory state.

Step 5: Deploy and Publish​

  • Click "Deploy" to publish the agent to the organisation.
  • Set access permissions for your team or organization.

Try It Yourself​

  • Experiment with different data sources and skills.
  • Orchestrate multi-agent workflows for complex tasks.
  • Monitor agent runs in Observability and apply security policies as needed.

For more advanced demos, see the Create Multi Agent Orchestrator and Integrations.