Classify Contract Types

In this example, we’ll be classifying various contracts by type using LlamaClassify. When provided with a file, we’ll classify it as either a co_branding contract or an affiliate_agreement.

We’ll use the CUAD dataset, which includes various contract types: Affiliate Agreements, Co-Branding contracts, Franchise contracts, and more.

The same workflow can be replicated in the LlamaCloud UI, where you can create classification rules visually instead of in code.

Install

Python

pip install llama-cloud>=1.6

TypeScript

npm install @llamaindex/llama-cloud

Download Contracts

Download a few contracts from the Affiliate Agreements and Co-Branding sections of the CUAD dataset.

Connect to LlamaCloud

Make sure you have a LlamaCloud API key. Set it as an environment variable or pass it directly to the SDK.

Initialize a Client

Python

import os
from llama_cloud import LlamaCloud

client = LlamaCloud(api_key=os.environ["LLAMA_CLOUD_API_KEY"])

TypeScript

import LlamaCloud from "@llamaindex/llama-cloud";

const client = new LlamaCloud({
  apiKey: process.env.LLAMA_CLOUD_API_KEY,
});

Define Classification Rules

For this example we’ll classify between co_branding and affiliate_agreements:

Python

rules = [
  {
    "type": "affiliate_agreements",
    "description": "This is a contract that outlines an affiliate agreement",
  },
  {
    "type": "co_branding",
    "description": "This is a contract that outlines a co-branding deal/agreement",
  },
]

TypeScript

const rules = [
  {
    type: 'affiliate_agreements',
    description: 'This is a contract that outlines an affiliate agreement',
  },
  {
    type: 'co_branding',
    description: 'This is a contract that outlines a co-branding deal/agreement',
  },
];

Classify Files

Upload a file and classify it:

Python

import time

file_obj = client.files.create(
    file="CybergyHoldingsInc_Affliate Agreement.pdf",
    purpose="classify",
)

# Create a classify job
job = client.classify.create(
    file_id=file_obj.id,
    configuration={"rules": rules, "mode": "FAST"},  # or "MULTIMODAL"
)

# Poll until complete
status = client.classify.get(job.id)
while status.status == "PENDING":
    time.sleep(2)
    status = client.classify.get(job.id)

if status.result:
    print("Classification Result: " + status.result.type)
    print("Classification Reason: " + status.result.reasoning)

TypeScript

import fs from 'fs';

const fileObj = await client.files.create({
    file: fs.createReadStream('CybergyHoldingsInc_Affliate Agreement.pdf'),
    purpose: "classify",
});

// Create a classify job
let job = await client.classify.create({
  file_id: fileObj.id,
  configuration: { rules, mode: 'FAST' },  // or 'MULTIMODAL'
});

// Poll until complete
while (job.status === 'PENDING') {
  await new Promise((r) => setTimeout(r, 2000));
  job = await client.classify.get(job.id);
}

if (job.result) {
    console.log("Classification Result: " + job.result.type);
    console.log("Classification Reason: " + job.result.reasoning);
}

Once classification is complete, the results include the predicted contract type and the model’s reasoning. For example:

Classification Result: affiliate_agreements
Classification Reason: The document is titled 'MARKETING AFFILIATE AGREEMENT' and repeatedly refers to one party as the 'Marketing Affiliate.' The agreement outlines the rights and obligations of the 'Marketing Affiliate' (MA) to market, sell, and support certain technology products, and details the relationship between the company and the affiliate. There is no mention of joint branding, shared trademarks, or collaborative marketing under both parties' brands, which would be indicative of a co-branding agreement. The content is entirely consistent with an affiliate agreement, where one party (the affiliate) is authorized to market and sell the products of another company, rather than a co-branding arrangement. Therefore, the best match is 'affiliate_agreements' with very high confidence.