Legal
Act Legal
M&A purchase agreements auto-filled from Chamber of Commerce data
Benchmarking Legora OS against an in-house Claude Code plugin on the same test deal, with an analyzer report breaking accuracy down per data field. Target: from 2–4 hours to 10–15 minutes per agreement.
Results achieved
Targeting 80% time savings per deal
Legora OS ~94% accuracy
In-house plugin ~89% accuracy
The challenge
The M&A team adjusts purchase agreements per deal in dozens of places — party data, directors, registered offices, Chamber of Commerce numbers and corporate cascades. That work currently costs 2–4 hours per agreement in find/replace rather than legal review.
Our approach
- 1.Iterative testing against one fixed benchmark deal
- 2.Benchmarking Legora OS and an in-house Claude Code plugin side by side
- 3.Analyzer report per data field and contract section (correct/missed/incorrect)
- 4.Next.js web app to compare runs visually
The solution
An evidence-backed decision track that shows which route — SaaS, in-house plugin or combination — is the best fit for act legal, with KvK always leading for factual data.
Claude CodeLegora OSPythonNext.jsKvK API