Ir para o conteúdo

How to build, expose, and embed AI agents with Jitterbit Harmony

Introduction

Jitterbit's Harmony platform empowers developers to build enterprise-grade AI agents, expose them as APIs, and embed them into intelligent applications using your choice of either low-code or no-code methods. This is accomplished through three primary functions within the platform:

  1. Build an AI agent as an integration using Integration Studio.
  2. Expose that AI agent as an API through API Manager or the Jitterbit APIM AI Assistant.
  3. Embed that API into an application in App Builder or using the App Builder AI Assistant.

What is a Jitterbit AI agent?

A Jitterbit AI agent is an autonomous, decision-making AI system built on the Harmony platform. Unlike a traditional, linear workflow, an AI agent is capable of performing the following actions:

  • Autonomously decide the best way to accomplish a given objective.
  • Independently determine the optimal methods and tools to achieve a goal.
  • Adapt and strategize to complete complex, multi-step tasks that may involve reasoning and planning.

Using this documentation

The Jitterbit AI documentation on building AI agents is divided into three main parts:

  • Reference architecture for building advanced agents: Start here to understand the "why" of Jitterbit AI agent design components. This section explains the high-level architecture, key components, and the end goal of a powerful AI agent.

  • Best practices: You should become familiar with our recommended best practices for Jitterbit AI agent design to keep in mind while building AI agents.

  • How-to guides: These are the step-by-step guides for the "how" of Jitterbit AI agent design. We offer a progressive path to build your first Retrieval-Augmented Generation (RAG) agent from scratch, and additional standalone guides:

    • Build a reactive AI agent: Start building a basic AI agent that responds to user queries using a large language model (LLM) without memory or advanced tools. This how-to also includes exposing the AI agent as an API with API Manager.
    • Build a contextual AI agent: Add on to the reactive AI agent by adding memory and context. This agent stores conversation history and maintains context across multiple interactions.
    • Build an AI agent with RAG: Add on to the contextual AI agent by adding a tool to address specific questions. This agent uses the Retrieval-Augmented Generation (RAG) technique, which combines LLM reasoning with access to external tools and data sources.
    • Build an AI agent with MCP: A separate guide shows how to build a basic AI agent that leverages the Model Context Protocol (MCP) to execute tools on an MCP server and act as a chatbot assistant.
    • Build an intelligent app: This separate guide is to be followed after you have created an AI agent and exposed it as an API with API Manager. Once you have exposed the AI agent, you can embed it as an API in an App Builder application to create an intelligent application.

Getting started

The fastest way to begin building an AI agent is by leveraging the existing assets in Jitterbit Marketplace:

  1. Log in to the Harmony portal at https://login.jitterbit.com and open Marketplace.

  2. In the Filters pane under Type, select AI Agent. The available AI agents are displayed.

  3. Click the AI agent's Documentation link to open its documentation in a separate tab. Keep the tab open to refer back to after starting the project.

  4. Click Start Project to open a configuration dialog to download customizations files (if applicable) and create the AI agent as an Integration Studio project.

  5. After clicking Start Project, complete the configuration dialog(s), then click Create Project.

  6. Once the progress dialog indicates the project is created, use the dialog link Go to Integration Studio or open the project directly from the Integration Studio Projects page.

  7. Follow the instructions to configure the AI agent provided via the Documentation link or accessed directly on this documentation website. Jitterbit AI agent documentation can be found under Jitterbit AI agents in Jitterbit Marketplace.

  8. Customize the AI agent for your automation needs using the information and how-to guides provided within the documentation.