Welcome to

watsonx Flows Engine

Build ready-to-use generative AI flows with the capability for customization and creation of new flows through an intuitive flow language, application developers and data scientists are equipped to easily construct AI applications that are both secure and high-performing.

Getting Started with RAG example

Try an end-to-end example utilizing RAG to augment your LLM responses through the context of your very own data.

Documentation

Detailed instructions to get you up to speed on how to make the most of watsonx Flows Engine.

Join the Discord community

Get help on using watsonx Flows Engine, or discuss AI-related topics.

Benefits

Empower developers of all skill levels

Accelerate concept to deployment

Built with evaluation, debugging, and monitoring capabilities, flows engine in watsonx.ai is an all-in-one developer tool to help bring generative AI applications from Proof of Concept to production quickly.

Maximize the value of your AI investments

Accelerate the realization of AI benefits for businesses by providing developers with comprehensive tools to enhance application quality and reduce deployment timelines, ensuring a quicker return on AI investments.

Key features

Create a GenAI app using a CLI and SDK with pre-built building blocks or custom steps. Use cases include RAG and function calling using API Schemas.

Pre-formated set up

Leverage out-of-the-box, high-performance generative AI flows that effortlessly connect with your enterprise data and business environment.

Flows Engine connections

Select from the options above to get started

To explore watsonx Flows Engine you get access to models from watsonx.ai without having to create your own watsonx.ai instance.
Note: You need to have a watsonx.ai instance to use Flows Engine in a production environment. Sign in to watsonx.ai to get keys for your instance.
Fair Use
In exceptional circumstances it may be necessary for IBM to take steps to suspend or limit unreasonable use by Licensee of the Preview Cloud Service, as described in the Service Description.