Quick Start

Upon logging in, you’ll be greeted with a helpful walkthrough modal, guiding you through the application’s features. You can choose to follow the step-by-step explanation of the UI or skip it entirely.

Once you’ve finished the walkthrough or decided to skip it, you’ll see a dropdown menu. From there, you can choose to either select a built-in template or start a new project from scratch as shown below:

We currently offer two standard templates:

Document QA: Chat with an LLM that understands your documents
Smart Bot: Chat with an LLM that has access to the internet

Choose a template and select Import If you prefer to start from scratch, select Start New Project

Main Page

Once you’ve started clean or imported your template you will be taken to the main page. The main page is split into 4 sections: Project Files on the left, Canvas in the middle, Action Palette to the right, and Event Logs at the bottom.

Dark Mode

You can seamlessly transition between Light and Dark themes by utilizing the Toggle Theme button positioned conveniently above the Action Palette. This feature offers you the flexibility to customize your visual experience according to your preference.

Project Files

Upload files from your computer or from a storage bucket to be used in your pipelines. Visit the Project Files page to learn more.

Canvas

The canvas is the space where your pipelines come to life. Pull components from the palettes onto the canvas, drag, rearrange, and connect the blocks to create your pipeline. Once you are finished, run the pipeline using the run button in the top right.

Action Palette

The action palette contains the blocks to build your pipelines. There are two types of palettes:

LLM Action Palette: Harness the power of LLMs to automate processes or generate content
NLP Action Palette: Assemble powerful NLP pipelines to clean, manipulate, analyze, and visualize text and tabular datasets

At this point in time, the palettes cannot be mix and matched. Each palette can create their own pipelines, respectively.

Event Logs

Event Logs will give you a window into your pipeline runs. You will see the messages output from your runs as well as any errors you may need to address. Learn more on the event logs page.