
Before you can start building AI applications using watsonx.ai Flows Engine:

**1. Sign up** for a free account

**2. Download & install the [CLI](/docs/installation)** for Python or Node.js

**3. [Authenticate](/docs/authentication)** your account

You can then start building your own tools or follow one of the example application tutorials below.

<br/>

### Learn how to build tools

Building a tool from your own data source consists of three steps:

**1. Build a tool**

- By importing your own data source (REST, GraphQL, Databases) [[Python]](/docs/python-cli/build-tools) [[Node.js]](/docs/node-cli/build-tools)
- Importing a prebuilt tool [[Python]](/docs/python-cli/import-tools) [[Node.js]](/docs/node-cli/import-tools)

**2. Register your tool(s)**

**3. Deploy** a watsonx.ai Flows Engine endpoint with your tool definitions

<br/>

The deployed endpoint can then be used in agentic frameworks such as LangGraph. See [integrations](#) for more information.

<br/>

### Build example applications

You can also start building by using one of our many tutorials:

- [Build a Tool Calling Agent](https://developer.ibm.com/tutorials/awb-build-tool-calling-agents-with-langgraph-and-flows-engine/)
- [Build a RAG application](https://developer.ibm.com/tutorials/awb-build-rag-application-watsonx-ai-flows-engine/)
- [Build a web page summarization tool](https://developer.ibm.com/tutorials/awb-build-web-page-summarization-tool-watsonx-ai-flows-engine/)
- [Implement LLM guardrails for RAG applications](https://developer.ibm.com/tutorials/awb-how-to-implement-llm-guardrails-for-rag-applications/)
- [Using different LLMs in watsonx.ai Flows Engine](https://developer.ibm.com/tutorials/awb-using-different-llms-in-watsonx-ai-flows-engine/)
  - [IBM Granite](https://developer.ibm.com/tutorials/awb-using-ibm-granite-llms-in-watsonx-ai-flows-engine/)
  - [Meta Llama 3.1](https://developer.ibm.com/tutorials/awb-using-meta-llama-3-1-in-watsonx-ai-flows-engine/)
  - [Mistral Large](https://developer.ibm.com/tutorials/awb-using-mistral-ai-llms-in-watsonx-ai-flows-engine/)

More tutorials will be coming soon. Requests for specific topics? Please let us know on [Discord](https://ibm.biz/wxflows-discord).


Learn how to initialize new projects using the watsonx.ai Flows Engine CLI on your local machine.

Getting Started

Overview

Welcome watsonx.ai Flows Engine

Installation

Install watsonx.ai Flows Engine

Authentication

Authenticate to the CLI

Build a Tool

Build a Tool from Your Data Source

Import a Tool

Import a Community Tool

Register Tools

Register Your Tools

Deploying Tools

Deploying Your Tools to an Endpoint

Python CLI

Using the Python CLI

Node.js CLI

Using the Node.js CLI

Using the CLI

Learn how to use the CLI

Building flows

Building flows using the flow language

Connect Vector Databases

Connect to different Vector Databases

Consume flows

Consume flows using a SDK

Connect to watsonx.ai

Learn

Learn how to use watsonx.ai Flows Engine

completion

contextFromTopN

graniteChatPrompt

hallucinationScore

promptFromTopN

ragAnswerInput

ragInfo

ragScoreInfo

ragScoreMessage

templatedPrompt

topNDocs

wx_promptLabCompletion

Flows engine steps

Learn what steps are available to start building flows

enum AIEngine

input GenParametersInput

interface HallucinationScore

scalar JSON

type MessageOutput

type Prompt

type PromptContext

type RagAnswerInput

type ScoreInput

enum SearchEngine

interface TopNDocs

interface _Normalize

type _Normalized

Types