Contract Optimization Using AI Tech
Budget: $50 – $0 USD
AI-Driven Contract Analysis and Management Platform
Task definition
Jobs to be done
As a Commercial Management Professional, I want to be able to:
Get a summary of the key contract terms (one A4 page to be automatically generated that covers all the Ts&Cs, liabilities, indemnities, LDs, notices, etc.)
Get advice on whether the contract is agreeable and terms are in line with what is normally expected:
LOGIC Contract
FIDIC (Silver, Red, Yellow, Silver, etc.)
BIMCO (Charter Party for Vessels and barges)
Charter Party for helicopters
Most difficult (Bespoke contract)
Get an indication of points that have to be consulted about with the lawyer:
To be trained by Lawyers
Get a proposal on how the contract should be reworded to satisfy internal requirements.
“Teach” the system what is right and what is wrong according to my company's standards and users' recommendations (e.g., which clauses can be used for a rejection or approval, Intensity in terms of Order of Precedence (which clause takes over the other one), which clause can dismiss the claim, etc.
Allow a toggle to decide how a paragraph or position is detailed using the different clauses of the Contract. The user can decide whether to use the accept or reject clauses at their disposal. The user should be in control of all the statements used in their replies.
Definition of inputs
From the user standpoint:
User uploads a contract
User can possibly provide some additional information / guidance / documents specific to their company (but this might not be for MVP as we will need to segregate info provided by different customers and have different models for them)
From the platform standpoint:
We will need to build some training sample based on existing contracts and, ideally, changes that have been made to them; it might possible to use GenAI to classify and process it by itself but we might also do some sort of manual mapping to highlight key points that shall be
Definition of outputs
Summary of the key contract terms
Advice on whether the contract is agreeable and terms are in line with what is normally expected
An indication of points that have to be consulted about with the lawyer
A proposal on how the contract should be reworded to satisfy internal requirements
ChatBOT to answer questions, give recommendations, alter responses and refer to clauses
Give fair recommendation based on the pre-set standards however, also proposing alternative wording and positions with caveats - Ultimately, giving the user full control and ownership of the data and answers!
What does success look like
The prototype is able to consistently (95+% of cases) identify all relevant things and propose reasonable changes
The prototype is able to evolve based on users / background input (e.g. free-form feedback from the user, additional documents, definition of expected outputs etc.)
The prototype is built and operates at minimum cost and it is clear how to evolve it further
The prototype generates actual value for potential customers
The prototype learns from its user and re-issues drafts and answers in line with the user’s preference.
User friendly process (Grammarly-like recommendation and interface)
Task definition
Jobs to be done
As a Commercial Management Professional, I want to be able to:
Get a summary of the key contract terms (one A4 page to be automatically generated that covers all the Ts&Cs, liabilities, indemnities, LDs, notices, etc.)
Get advice on whether the contract is agreeable and terms are in line with what is normally expected:
LOGIC Contract
FIDIC (Silver, Red, Yellow, Silver, etc.)
BIMCO (Charter Party for Vessels and barges)
Charter Party for helicopters
Most difficult (Bespoke contract)
Get an indication of points that have to be consulted about with the lawyer:
To be trained by Lawyers
Get a proposal on how the contract should be reworded to satisfy internal requirements.
“Teach” the system what is right and what is wrong according to my company's standards and users' recommendations (e.g., which clauses can be used for a rejection or approval, Intensity in terms of Order of Precedence (which clause takes over the other one), which clause can dismiss the claim, etc.
Allow a toggle to decide how a paragraph or position is detailed using the different clauses of the Contract. The user can decide whether to use the accept or reject clauses at their disposal. The user should be in control of all the statements used in their replies.
Definition of inputs
From the user standpoint:
User uploads a contract
User can possibly provide some additional information / guidance / documents specific to their company (but this might not be for MVP as we will need to segregate info provided by different customers and have different models for them)
From the platform standpoint:
We will need to build some training sample based on existing contracts and, ideally, changes that have been made to them; it might possible to use GenAI to classify and process it by itself but we might also do some sort of manual mapping to highlight key points that shall be
Definition of outputs
Summary of the key contract terms
Advice on whether the contract is agreeable and terms are in line with what is normally expected
An indication of points that have to be consulted about with the lawyer
A proposal on how the contract should be reworded to satisfy internal requirements
ChatBOT to answer questions, give recommendations, alter responses and refer to clauses
Give fair recommendation based on the pre-set standards however, also proposing alternative wording and positions with caveats - Ultimately, giving the user full control and ownership of the data and answers!
What does success look like
The prototype is able to consistently (95+% of cases) identify all relevant things and propose reasonable changes
The prototype is able to evolve based on users / background input (e.g. free-form feedback from the user, additional documents, definition of expected outputs etc.)
The prototype is built and operates at minimum cost and it is clear how to evolve it further
The prototype generates actual value for potential customers
The prototype learns from its user and re-issues drafts and answers in line with the user’s preference.
User friendly process (Grammarly-like recommendation and interface)