Workflows
Last updated
Last updated
Workflows are the core of the Lleverage platform. They enable product and engineering teams to build and manage AI features by chaining together business logic, data, APIs, and dynamic prompts. Each workflow can be exposed as a single API endpoint, making it easy to integrate AI functionality directly into your applications.
Workflow Studio: The Workflow Studio is a visual tool where teams design, build, and refine their AI workflows. It allows you to connect different components, known as Nodes, to create complex chains of functionality. Nodes represent self-contained blocks like data processing, API calls, or generative AI prompts, and they can be linked together to form a cohesive AI feature.
Workflow Engine: The Workflow Engine is responsible for running the workflows created in the Studio. It's the invisible part that executes the workflows when triggered by your application. Every Lleverage organisation runs a dedicated instance of the workflow engine. Your own private, auto-scalable cluster, where only your workflows run and secrets can be accessed.
Workflows in Lleverage can handle a wide range of use cases, from automating processes to integrating with external services. They give you the flexibility to define how data flows, how APIs interact, and how prompts are optimized to create powerful AI-driven functionality.
By using workflows, you can manage the full lifecycle of AI features—from development to production—while keeping everything organized within your Project.
Lleverage supports a wide range of file types, allowing you to work with various forms of text, data, and images within your workflows. This flexibility makes it easy to incorporate different content types and even enables the use of multi-modal models for processing images alongside text data.
Supported File Types:
Text and Data-Based Files:
PDF: Upload PDF documents for text extraction, processing, and embedding.
PowerPoint (PPTX): Use PowerPoint presentations as input for workflows, enabling slide-by-slide analysis and embedding.
Word Documents (DOCX): Process Word documents for natural language tasks, embedding, or extraction.
Excel (XLSX): Handle Excel files, making it possible to analyze tabular data or incorporate structured data into workflows.
CSV: Easily upload CSV files for structured data processing, analysis, or bulk testing in workflows.
Images:
Image Files: Supports standard image formats (e.g., JPEG, PNG) for processing with multi-modal models. This enables workflows that analyze both visual and text-based data, opening up possibilities for use cases like image captioning, visual recognition, or combined text-image analysis.
These supported file types allow you to leverage different data sources seamlessly in your workflows, whether it’s processing structured data, analyzing documents, or working with image