<aside> 💡

Problem:

Legal users need to write while simultaneously loading a huge context window. They will be constantly looking into prior contracts, which may be scattered across many places.

</aside>

<aside> 🧱

Current flow:

• A lawyer currently receives a contract via email and access them via editing tools such as Google Docs or Microsoft Excel. In rare occasions, legal professionals will use tools such as Pages or alternative text editors.
• When drafting a contract they must understand several categories of legal concepts in the contract. They will need to check if the contract can be compared to prior cases or company policy. If the clauses are acceptable or not and if they have used similar language before.
• A lawyer will review prior cases manually or using a traditional search engine. In some cases, they will need access to specialized databases. In the case of documents used before (or precedents) they will juggle multiple docx or similar format files.
• They will incur spending a lot of time formatting documents to meet the formatting of the case. Contracts may have different styles for header based numbering. </aside>

<aside> 🚥

Solution hypothesis:

The hypothesis is that lawyers need a text editor that is an assistant in context . An AI agent may present itself as a text editor and not necessarily as a layer on top the current legal review workflow. The editor should be able to make drafting recommendations right there in the text editing tool and those responses can be augmented by the user’s own or selected knowledge base. It can quickly advise on wether to use prior language or make modifications to the language so that it matches company policy.

</aside>

<aside> 🕰️

Business impact:

The impact of faster contracting is two folded. It can allow for cost reduction because lawyers will be more productive and will review more contracts in less time. Additionally, it can help make revenue closing cycles shorter.

</aside>

<aside> 📣

Size estimation:

We want to start with a minimal viable product that is a simple text editor but that is AI native. The initial objective will be to build a solid AI architecture of agents that can generate responses based on the given knowledge and suggest changes to the document under review. One key feature is giving access to prior uses of similar language and that is able to provide references for each quote.

</aside>