Main / Products / Contract Analysis

Contract Analysis

AI Contract Analysis transforms a long, tedious process into an efficient instrument for your business, saving resources, enhancing workflows, and lowering risk. A mix of hand-crafted rules and Machine Learning helped Silk Data launch a web interface for businesses to copy, and paste, and analyze contracts. Contract Analysis is the automated evaluation of legal terms related to a specific transaction or business relationship.

Challenge

Almost every business deals with contract preparation, review, and signing. Usually, having a manual contract analysis process in a company demands a lot of time, money, and attention. Furthermore, there are fewer specialists who are ready to spend resources on it. In the legal community, contract analysis is widely considered a boring and complicated procedure. The global purpose was to automate this process and save resources by doing that.

Solution

Working with our solution is straightforward for the end user, while the algorithm behind the scenes is rather complex.

The user can paste the content of a contract or upload a PDF or DOCX document, and in just a few seconds, AI will show key positions in the contract. That Contract Analysis speeds up productivity by cutting down the procedure from hours to minutes.

Let's discuss your next project together!

Purposes

To make the process as convenient as possible, we focus on specific goals:

Create a kind of automatic assistant or a ‘spell-checker’ for legal experts. This product is intended for both experienced professionals and junior specialists. Professionals can relieve the stress of monotonous work by optimizing this process. While the main benefit for beginners is practice in seeing key points in contracts.
Fast onboarding. We wanted to create a ready-made solution for small or medium-sized businesses.
Flexibility. Support of different languages and legislatures.

Our approach

It's important because the contract type defines applicable law provisions and contract negotiation playbook use. Supported types of contracts in the first version:

Purchase contracts
Consignment (supply) contracts
Service contracts

Steps

Step 1. Contract classification

To determine the type of contract accurately, Silk Data uses a mixed approach. Firstly, we start with the contract and analyze it with Machine Learning model. This step defines if it is a service contract or not. The next step is where hand-crafted rules appear. With the help of specific patterns, it can tell the slight difference between supply and purchase contracts. A mixed approach with Machine Learning model and hand-crafted rules allows to achieve better performance results. It ensures high classification accuracy, ca. 90%.

Step 2. Clauses classification

After determining the contract type, models of legal concepts and document segmentation find important clauses in the contract text.

Step 3. Legislation

The legislation purpose is to confirm clauses with the expert-provided links. Currently, we are developing automated suggestions for law provisions and edits.

Results

The primary outcome is provided web interface makes it easy for business to upload and analyse DOCX and PDF contracts. You can test our MVP here.

Silk Data is also developing a special MS Word add-on to bring the Contract Analysis directly to the typical tool of an office worker. The idea of the extension is to make minimal intrusion into the document, only selecting the important provisions. Processing of DOCX and PDF files can be added to the web interface on demand.