Skip to content

Introduction to Studio

Studio is where projects and ideas come to life. Think of Studio as the heart of INTELLITHING—it's where you'll be spending most of your time. While there are key features like training, data generation, etc., Studio is where you stitch all those individually created components together.

Key Concepts

Studio has two main sections: the Block Editor and Workflow. In the left-hand side navigation, you'll find:

1. Deployment and Utilities

After creating and saving your project, Deployment and Utilities is where you'll find critical operations related to deployment and testing, including:

  • Action Button: This button allows you to build a Docker image and deploy it. It also provides operations such as stopping the deployment, deleting the project, and migrating the Docker image to Docker Hub.

  • Status: The project status bar shows two different states: Registration and Deployment.

    • When you save a project, it is automatically registered—this means the project dependencies are downloaded.
    • Once registered, you can deploy it. During deployment, you'll see three stages: building, pushing, and deploying, followed by an initialisation stage at the end.
    • This means INTELLITHING is building a Docker image for your project, pushing it to your registry, and finally deploying it on Kubernetes. "Initialising" means the system is waiting for the deployment to run, and logs will be available from this point onward.
    • When the status turns to Deployed, it means the project has been successfully deployed.

  • Project Card: Here, you can modify the project name and description or add collaborators to the project.

  • Deployment Logs: At the bottom of the page, you can see your deployment logs, which help you troubleshoot deployment issues.

  • Inference API / Deployment API: Once the deployment is successful, the inference API is automatically generated and populated. You can copy it from the Deployment API section.

  • Chat Now: Chat Now is primarily used for applications triggered by chat or powered by conversational interfaces. It's a testing/playground feature available immediately after a successful deployment. When you click Chat Now, the API endpoint is automatically retrieved, and a chat UI opens in a new window—only accessible to you. This allows you to test the pipeline in action and view logs for visibility and clarity.

2. Security

Security is where you can create and manage bearer tokens for your deployment API, and configure other security-related settings—such as whether you want to make the deployment publicly accessible as a chat interface for external users. Your API endpoint is always separately available, but if you decide to launch the project as a public chat, a UI similar to the one in Deployment and Utilities becomes available for sharing with people outside your organization.

Within the Security section, you’ll find:

  • Create a New Key: Generates a bearer token for the API.
  • Publish Publicly: Enables a shareable chat UI for people outside your organization.

3. Analytics

This section gives you full visibility into the performance of your project—from the deployment pipeline to cloud and hardware usage for this specific project.

Key Definitions

1. Block Editor

Blocks are drag-and-drop modules that can be individually configured and manipulated. This process is done through a visual editor called the Block Editor. A combination of blocks is referred to as a stack. Every block consists of a combination of nodes.

  • Combination of nodes → Block
  • Combination of blocks → Stack

Within the Block Editor, you have:

  • LLMs: Mostly open-source, Hugging Face-compatible models.
  • Data Connectors: These connect to various data sources to retrieve data dynamically during pipeline execution.
  • Agents: Includes React Agents and Function Calling Agents.
  • Supervised Machine Learning: Models trained via the “Train a Model” section, but available in the Block Editor like other modules.
  • INTELLITHING Router Engine: Every stack uses the INTELLITHING Router Engine by default to route queries to the appropriate block. However, workflows allow you to override this default behavior and create a custom flow.

2. Workflow

The Workflow defines the flow of actions, data, inputs & outputs, relationships, and behaviors within a stack. Once you create a stack, a default workflow is automatically generated. You can then customize it to match your specific process and create a highly personalized application.

The need for workflow arises because the default behavior of a stack cannot cover all possible use cases. By default, an LLM is central to any stack, with all blocks orbiting it. When you send a query to the LLM, it is routed to a system selected block with limited inpu&output manipulation. Hence workflow is to overcome this.

Within the Workflow, you have:

  • Nodes: A block is a combination of nodes or steps needed to complete a task. These nodes are visible in the workflow but not editable—they represent the core logic and are designed to be dynamic, versatile, and powerful using state-of-the-art techniques.

  • Bridge: A Bridge is a core feature used to modify workflows. It connects two blocks together. Once a bridge is added, the stack will use your custom configuration instead of the INTELLITHING Router Engine. You can add as many bridges as you need, but a bridge can only connect the last node of one block to the first node of another. In other words, it transfers the output of one block to the input of another.