We are excited to report on the launch of a new open source software project by advisory board member Shawn Curran and his law firm, Travers Smith. The software product is called Etatonna and it uses AI to help identify and label concepts and clauses in contracts for the purpose of automated review.
Shawn is the Head of Legal Technology at Travers Smith, a pioneer in developing open source software for the legal market. We wrote about his earlier open source software project called MatMail in October of 2019. Travers Smith is a corporate law firm headquartered in London, United Kingdom. It advises national and multinational companies in the UK and internationally across the full range of corporate and commercial matters.
The goal for Etatonna is to provide a system that will allow lawyers and clients to convert contracts, primarily unstructured data, into reusable structured data. The software allows law firms or clients to upload contracts into the system and then flag key clauses for both analysis and reuse. As an example, consider the standard “force majeure” clauses included in many contracts. Etatonna allows a reviewer to highlight and identify the contract, label it by type and even indicate whether it is favorable to either side of the transaction. At a later time, the firm or its client could pull back examples of such contracts for further use.
The initial intent is to turn contract clauses into legal lego blocks, with the hope of improving the generation of contacts in the future. Once sufficient clauses are added to the system, a user could assemble parts of a new contract automatically using Etatonna’s structured information. Here is an example of clause labeling in the system:
Even more exciting, Travers Smith plans to go beyond document assembly to bring AI into the process. Think of the contract blocks as training materials for a machine learning algorithm. Once the algorithm is sufficiently trained, it could then analyze and identify clauses in new contracts fed into the system.
Open sourcing Etatonna seemed a brilliant, albeit commonsense, next step in the process. Law firms and clients are naturally concerned about the confidentiality of their data and even the structure and content of their contract clauses. By offering Etatonna into open source, Travers Smith provides a way for law firms to collaborate on legal efficiency while not risking privacy and their work product. Each can run their own version of the software and, at least logically, each could bring in their preferred AI tool to analyze the seed clauses.
A better step, in our judgment, would be to pool these resources through a neutral provider so that the cumulative value of review and training could go across clients and law firms to better train a central AI algorithm. Doing so might make the algorithm much smarter, enabling it to better label new contracts brought to its attention. Showing recommended force majeure clauses, or even suggesting alternatives would not seem to risk much in the way of privacy and would make better use of the combined resources of each of the sharing law firms. Should some be worried about a loss of work product, that has always been the case once the contract is placed into public commerce. For ages lawyers have been collecting good drafting examples and reusing other’s clauses. It’s just how things are done.
Here is a fun and creative look at the Travers team’s vision for Etatonna:
Ultimately, Traver’s desire was to give business clients full autonomy over what data exists within machine learning models allowing them to retain their investment in data and training and even share datasets to build joint venture-type AI models. They can retain lawyers to label contract provisions or recruit smart but non-legal reviewers to do the job at lower cost. As Travers Smith pointed out in its press release:
“During the development of Etatonna it became clear that certain legal documents lend themselves to being labelled by non-experts. For example, clear headings and short paragraphs, in almost all cases, will provide enough context for non-experts to label correctly. This means that users do not need to rely on legally trained experts in order to use the software efficiently.”
While it remains to be seen whether products like Etatonna will induce law firms to collaborate for great legal efficiency (bringing the costs of legal services down), we tip our hat to Shawn and the Travers Smith team for having the foresight to build software to improve their efficiency, for sharing these tools with their clients and for making them available to their “competitors” for the purpose of making the delivery of legal services more efficient.
Organisations that would like to access the open source code can register by emailing Shawn Curran at firstname.lastname@example.org. Etatonna is licenced under the GNU GPL v3 licence and is shared through the existing Travers Smith Open Source environment on Azure DevOps.