Clausecraft
Contracts drafted in hours, reviewed with total recall.
- Client
- Whitfield & Marsh LLP
- Year
- 2021
- Duration
- 7 months
- Team size
- 4 engineers
Interactive Demo
A hands-on mock of the product we built — click around.
Demo interface shown in English.
The challenge
Whitfield & Marsh’s corporate team turned around 1,300 commercial contracts a year, each first draft consuming three weeks of associate time and partner review. Knowledge lived in inboxes: the perfect indemnity clause drafted in 2019 existed somewhere, but finding it meant asking around, and every lawyer redrafted from memory with slightly different risk. Clients increasingly demanded fixed fees, squeezing margins on exactly the work that consumed the most hours, and a botched limitation-of-liability clause had already cost the firm a negligence claim. Any tool had to respect privilege absolutely — UK-hosted, access-controlled, with an audit trail a court would accept — or the partners would never let it near live matters.
Our solution
We built Clausecraft as a React workspace over Node.js services, with a Python NLP pipeline that classifies incoming contracts against the firm’s 140-type clause taxonomy, flags deviations from house-standard positions and scores risk per clause. A matter-aware Elasticsearch index makes fifteen years of precedents searchable in seconds, permissioned so conflict walls hold. Document assembly merges approved clauses into Word templates with tracked changes intact, and an e-signature flow closes the loop with a tamper-evident audit log of every view, edit and approval. Everything runs in UK data centres on Kubernetes, encrypted per matter, with SSO against the firm’s Active Directory.
Results
Technology stack
Facing a similar challenge?
We've solved this class of problem before. Let's talk about how that experience applies to your situation.
Discuss your projectNext case study
Forgepulse →