Skip to content

Mastercard Enhances Contract Analytics Using AI & Machine Learning Technology

Mastercard leverages AI and Machine Learning to enhance the quality of its contract metadata and analytics capabilities. This results in improvements to information access, collaboration, and service delivery across Legal in relation for customers and internal analysis.


The Challenge

Mastercard teamed with Mainspring to cleanse its existing CLM repository to share contract information and mitigate risk from engaging high-risk vendors, confirm presence of key contract provisions, and minimizing spend not under contract.  Mastercard’s CLM system was implemented several years ago and over time, experienced a degradation in its data integrity through poor user adoption. Executives supported the effort for a recalibration of the data and enhanced analytics capabilities while rebuilding its CLM platform.

Our Solution

Mainspring, working collaboratively with Mastercard, drew up a data entry model that would capture not just relevant fields for transactional information but additional fields that would enable enhanced analytics. Mainspring established a data cleansing schedule to meet internal deadlines while supporting the redesign of a new CLM program to ensure that future contracts collect necessary details during the creation process.  Lastly, Mainspring incorporated in its services using Kira Systems, a NLP machine learning system, to extract contract text from standard and non-standard agreements.

Mainspring delivered this effort for Mastercard in multiple delivery steps:

  1. The first step involves leveraging machine learning technology to OCR and extract contract provisions to meet specific Mastercard requirements.
  2. Mainspring’s contract analysts then manually review extractions to verify or make corrections which include transcribing handwritten text and establishing integrity of parent-child-sibling contract relationships.
  3. Verified data, measured at over 95% accurate, is then extracted from the technology to be transformed in a manner that can be ingested to Mastercard’s CLM system.  
  4. Using the saved models, the process can be reiterated as needed for additional batches of contracts.  

The Results

Mastercard’s current CLM repository is being used far more extensively - particularly to assess the quality of contracts with existing vendors while establishing contracts with new vendors. Mastercard also has ongoing access to the extraction technology to pull additional extractions on demand.


The application also shows insight into:

Causes & factors of
missing contracts

Timeliness & frequency of upcoming renewals and expiration dates

Categories & amount of spend not under contract

Executive support was critical.  Mastercard’s sponsorship of the initiative meant that it had to happen and deliver within its desired ambitious timeframe, with minimal burden and interruption to the business.

Trustworthy data via AI leads the way for turning information into insight.

Trustworthy data from applications saves stakeholders time and money, leading to increased user adoption and more effective companies.