LlamaAgents Builder: From Prompt to Deployed AI Agent in Minutes

In this article you will learn how to build, deploy, and test a no-code document processing AI agent using LlamaAgents Builder in LlamaCloud.

Topics we will cover include:

• How to create a document classification agent using a natural language vector.
• How to deploy the agent on a GitHub-supported application without writing code.
• How to test published proxy on invoices and contracts in LlamaCloud interface.

Let’s not waste any more time.

introduction

Creating an AI agent for tasks like analyzing and processing documents autonomously used to require hours of endless configuration, code formatting, and deployment battles. yet.

This article reveals the process of creating, deploying, and using an intelligent agent from scratch without writing a single line of code, using LlamaAgents Builder. Better yet, we will host it as an app in a software repository that will be 100% owned by us.

We’ll complete the entire process in a matter of minutes, so time is of the essence: let’s get started.

Build with LlamaAgents Builder

LlamaAgents Builder is one of the newest features in Llama Cloud Web platform, whose main product was originally introduced as Lamapars. I know, a bit of a confusing mix of names! For now, just keep in mind that we will be able to access the agent builder through it This link.

The first thing you should see is the main menu like the one shown in the screenshot below. If that’s not what you see, try clicking on the “LlamaParse” icon in the top left corner instead, and you should see that – at least at the time of writing.

LlamaParse main menu

Note that, in this example, we are working under a newly created Free plan account, which allows up to 10,000 pages to be processed.

Do you see the “Agents” block in the lower right side? This is where LlamaAgents Builder lives. Although it’s in beta at the time of writing, we can already create useful agent-based workflows, as we’ll see.

Once you click on it, it will open a new screen with a chat interface similar to Gemini, ChatGPT, and others. You’ll get several suggested courses of action for what you want your agent to do, but we’ll define our own by typing the following prompt into the input box below. Just natural language, no code at all:

Create an agent that classifies documents into “Contracts” and “Invoices.” For contracts, extract the signing parties; For invoices total amount and date.

Determine what the agent should do with a natural language prompt

Simply submit the claim and the magic begins. With a great level of transparency into the thought process, you’ll see the steps completed and the progress made so far:

AgentBuilder creates our agent workflow

After a few minutes, the creation process will be complete. Not only can you see the complete workflow diagram, which gradually evolves throughout the process, but you also receive a concise and clear description of how to use the newly created agent. Simply amazing.

Agent workflow created

The next step is Our agent posted So that it can be used. In the upper right corner, you may see “Push and publishThis starts the process of publishing your agent workflow software packages to the GitHub repository, so make sure you have a registered account on github Firstly. You can easily sign up using an existing Google or Microsoft account, for example. Once you connect the LlamaCloud platform to your GitHub account, it’s very easy to push and deploy your agent: just give it a name, decide if you want it in a private repository, and that’s it:

Push the agent workflow and publish it to GitHub

The process will take a few minutes, and you’ll quickly see a set of command line-like messages appear. Once it’s done and your agent status appears as “Run“, you will see some final messages similar to this:

The “Uvicorn” messages indicate that our agent is deployed and running as a microservices API within the LlamaCloud infrastructure. If you’re familiar with FastAPI endpoints, you might want to try them programmatically through the API, but in this tutorial, we’ll keep things simpler (we promised there wouldn’t be any coding, right?) and try everything ourselves in LlamaCloud’s own UI.

To do this, click “”VisitThe button that appears at the top:

The agent is deployed and running

Now comes the most exciting part. You should be taken to a playground page called “Review,” where you can try out your agent. Start by uploading a file, for example, a PDF document containing an invoice or contract. If you don’t have one, simply create your own mock document template using Microsoft Word, Google Docs, or a similar tool, like this one:

LlamaCloud Proxy Test UI: Invoice Processing

Once the document is uploaded, the agent starts working on its own, and within seconds, it will classify your document and extract the required data fields, depending on the document type. You can see this result on the right panel in the image above: The total amount and invoice date were extracted correctly by the agent.

How about uploading a sample document containing a contract now?

LlamaCloud Proxy Test UI: Node Processing

As expected, the document is now labeled as a contract, and on this occasion, the extracted information consists of the names of the signing parties.

I did well! As you continue to run examples, make sure to approve or reject them based on whether they have been processed correctly: this helps the agent learn from the feedback.

Agent test cases and their status

wrap

We’ve seen how to build and deploy an AI agent capable of classifying and processing documents in different ways depending on the document type, step by step and without any lines of code – all in just minutes and within LlamaCloud’s newly added feature, LlamaAgents Builder.