Components Overview
In Data Toolchain for AI, each node represents a functional component within your pipeline. A node can either bring data in (source), process or transform it, or store the output. Nodes can be connected together on the workflow canvas to build end-to-end data flows.
Aparavi is actively expanding the list of supported nodes.
Add and Configure a Component
To add a component, click on the + at the bottom right of the screen
which will open the Component Panel
Once you select the component, you can set the properties by clicking the GEAR icon. If you hover over the Component, you can also copy it or delete it.
You then can add another Component and link them together by matching the nodes (Text goes to Text, not Document, etc)
Component Types
Audio Components
Processes audio input β this may include transcription, format conversion, or embedding audio for ML usage.
Database Components
Connects to SQL or NoSQL databases, allowing for queries, lookups, or data enrichment within a pipeline.
Embedding Components
Transforms text or images into vector embeddings, enabling downstream AI models to work with structured representations of unstructured content.
Example – Embedding β Image.
Image Components
Handles image-specific tasks like resizing, format conversion, or preprocessing before embedding or analysis.
Infrastructure Components
Nodes for orchestration, automation, or triggering β potentially linked to Webhooks, schedules, or external systems.
LLM Components
Large Language Models used to generate, summarize, classify, or transform text based on context or input from previous nodes.
Preprocessor Components
Preprocessor components prepare data for further processing by cleaning, formatting, and transforming it.
Source Components
Connects to external systems and data repositories to bring raw or unstructured data into the workflow.
Examples – Google Drive, AWS S3, Outlook, SMB/NAS, Local File System.
Store Components
Used to write and persist processed output to a destination, such as cloud buckets, local drives, or database tables.Β
Text Components
Text manipulation utilities such as extraction, formatting, cleansing, or splitting operations within the workflow.
Other Components
Utility nodes that help manage data structures, formats, and flow β such as joins, filters, or conditional logic.
