New-To-World Feature Set Incorporating ML/AI


Evisort is a SaaS software product that offers contract management capabilities to organizations ranging from SMB’s to enterprises. 

It allows users to manage the entire lifecycle of contracts, from creation and negotiation to execution, storage, and analysis.

It uses AI to: 

  • Automate contract analysis
  • Streamline contract creation
  • Assess risks
  • Monitor performance
  • Provide renewal recommendations

The Challenge

One key capability that Evisort needed to add in order to maintain its market position was something called a “clause library.” A clause is a distinct provision, article, section, or paragraph within a legal document that addresses a specific point or aspect. 

Each section or paragraph of a legal document may contain multiple clauses.  These clauses serve to define rights, obligations, responsibilities, conditions, or other relevant aspects of the legal document.

An average legal agreement has about twenty clause types. Different types of agreements will have more or fewer clause types depending on the nature or purpose of the agreement. Furthermore, different agreement types will use different language in the same clause. 

Here are examples of a Termination clause from two different agreement types:

Service Agreement

This Agreement may be terminated by either party in the event of a material breach by the other party, provided that the non-breaching party provides written notice specifying the breach and allowing the breaching party a cure period of [X] days to remedy the breach. If the breach remains uncured after the cure period, the non-breaching party may terminate this Agreement immediately upon written notice, etc.

Consulting Agreement

Either party may terminate this Agreement for any reason upon [X] days’ written notice to the other party. In the event of termination, the terminating party shall be relieved of any further obligations, or responsibilities under this Agreement, except for those obligations that, by their nature, survive termination. Upon termination, all fees and expenses owed by the client to the consultant shall be paid within [X] days.

Since an organization can have many in-force agreements at a given point in time, unintended language differences across agreements pose significant information integrity and risk management problems. It increases the odds that an organization will become party to agreement provisions that they did not expect or desire.

Problem Statement

Customer organizations had no way to import, create, maintain, update, delete, and report against the clauses that comprised the legal agreements they entered into. 

So the opportunity was to provide better clause language management capabilities for counsel and allied legal operations staff.

Our starting problem statement was the following:

How might we…

  • Provide legal counsel and legal operations contributors with a way to organize and standardize their clause language in a single place, 
  • Define when and in what situations different clause versions should be used, 
  • Use the standardized language to construct legal agreements, 
  • Update or retire clause language as needed, and
  • Track clause language usage, 

So that time to agreement and use of non-standard clause language can be reduced?

My Approach

Working closely with a small team of one product manager, a visual designer, and one developer, I…

  • Organized and ran a discovery sprint to discover the key jobs that people needed to do to maintain and track their clause language.
  • Identified the target personas we were designing for.
  • Created and rapidly iterated a design strategy that would allow users to easily add clause language via uploading content or by leveraging the product’s AI to identify the most frequently-used clause types and language variations in their organizations’ corpus of agreements.
  • Ran multiple rounds of usability testing to validate our approach and fine-tune the design.


Between December of 2022 and May of 2023, the team and I:

  • Conducted an effective discovery sprint.
  • Shipped a barebones MVP two months after the end of the discovery sprint.
  • Gathered actionable feedback from MVP users.
  • Progressed from initial design strategy flows and sketches to prototype testing by March of 2023.
  • Conducted three rounds of usability testing and numerous design revisions between March and May of 2023.

The result: a de-risked final design and high confidence that the post-MVP release would meet users’ needs.

Want to see the full writeup for this case study? Contact me for access.