Henchman is a new NLP-driven Microsoft Word add-in that helps you to draft contracts faster by retrieving previously written clauses from your contract repository. After growing to ten staff and getting their product ready for market, the company has now raised €1m ($1.2m) and is formally launching after some time in stealth mode.
It’s based in Belgium, but is language agnostic – primarily because its NLP capabilities, which are used for hunting past clause examples, are running off your own work product, and that may be in many languages.
The €1m in funding is a mix of a Vlaio innovation grant, bank financing and private investors such as Alex Segers, (former CEO Domo Chemicals), and Tim Clauwaert, (former CEO Intuo).
Artificial Lawyer asked co-founders Wouter Van Respaille, Gilles Mattelin, and Jorn Vanysacker, (pictured L to R below), about the product.
– Why did you create this?
Henchman’s objective is to solve lawyers’ most annoying problem – the ‘I’ve written this before, but where?’ mantra. While drafting contracts, legal experts almost always rely on their own or their colleagues’ previously written contracts as a benchmark or template.
In the process of looking for a specific clause, or just general inspiration, lawyers open numerous folders and documents, making the search for previously written clauses very time-consuming – up to one hour per day is lost on searching for past files.
– How does this differ from other contract drafting tools already in the market?
Henchman does not plan to expand its product offering horizontally. We are determined to do one thing very well: Help lawyers draft contracts in the most user-centric manner, i.e. the fastest contract drafting experience ever made. Henchman is different mainly in its setup effort, which is zero and in stark contrast with solutions that require you to categorise your database.
Secondly, it stands out because of the ease of use: working within Word and getting suggestions in an intuitive search. There is no need to change the way of working and therefore the adoption rate is very much above average.
Our team consists of people that have SaaS experience and that allows us to focus and ‘put the user first knowing that all else will follow’. In doing so we aim to have 10,000 lawyers using Henchman by 2026.
– Please can you tell us more about the NLP side of the product?
We rely upon a wide array of NLP methods. Our search and analytics stack is built from the ground up to make use of the latest and greatest in machine learning, notably deep learning. The search engine itself is something of a hybrid of traditional information retrieval technology and neural representation learning methods.
Our models are custom-built and trained in-house on proprietary datasets that have been studied and enriched with human knowledge in close cooperation with a dedicated team of legal experts.
– Is this for English docs, or mainly French/Dutch?
Our technology is built language agnostic, meaning it works on any contract or document language in a lawyer’s database.
So, there you go. It looks to be a very handy tool. And it’s one that takes a complex problem and aims for a simple solution, i.e. to leverage your previous work product using NLP to help you to make your new contract.
Of course, the proof will be in the pudding, but it’s a solid idea that other legal tech companies are also working on, albeit all with differing approaches.
Fundamentally this is the kind of thing that really can save lawyers time (if it produces the right results…). There is something clearly so inefficient about remaking things that you already have that’s it great to see another legal tech startup try to solve this problem. Good luck to the team. (And if they get to 10,000 lawyers using the product within five years – as they hope – this site will definitely buy the team a drink!)