We are a machine learning company, specialising in applying ML to text. We can classify sections of text with high granularity and precision. We aim to be able to classify an individual contract clause to determine what kind of contract it came from and what kind of section it belongs to. For instance, imagine this clause: "The total consideration will be increased by 30% if the Company exceeds 15% in sales margin in the calendar year 2020". Our machine learning system should "understand" that is likely to be from the Earn-Out section of a Sale and Purchase agreement for the acquisition of a company.
That sounds complicated, and it is. However, for now we just want the "section skeleton" of a few types of contracts.
Take a "Sale and Purchase" agreement for the acquisition of a company as an example. It might contain sections like: Definitions and Interpretation, Jurisdiction, Defined Terms, Interpretation, Share purchase, Consideration, Warranties, Undertakings, Leakage, Earn-out, Confidentiality, Announcements, Liability, Guaranteed, Warranties, Indemnity, Notices, Governing law.
We would like that kind of skeleton of headings (but more detailed and comprehensive) for three different types of contracts:
1: Sale and Purchase Agreement (acquisition of a company or a business)
2: HR Termination agreements for individuals
3: Purchase of buildings
We would like to have as much detail as possible. If there is more detail then we would like the sub-sections to clearly show the parent section.
The deliverable is a list of section headings (ideally with a subset of headings under each) for each of the contract types.
You will need review sample contracts to do the research. If you do not have access to such, consider using contracts filed by publicly traded companies or contracts made public because one of the parties is a government body.
The contracts you base this on should be written for U.S. and English law