CestaLaw

Overview

Sister product to CestaLex, focused on European Court of Human Rights case law and expanding toward wider European jurisdictions.

• ECHR judgments have their own citation structure and procedural logic; the search and citation layer was modelled for this corpus specifically.
• Kafka handles inter-service communication and Redis caches the hot read paths.
• An AI microservice plugs into the drafting flow, with model calls kept off the request path.

01

Problem

02

Approach

03

Architecture

04

Key decisions

1. 01

   Share the infrastructure, not the domain model

   Identity, invoicing and the event backbone are shared; corpus-specific search and citation logic is not. Both products move at the same speed without corrupting each other's domain rules.

2. 02

   Put the hot read paths in Redis

   Judgment pages are read often and change rarely. The Redis cache takes load off the database and speeds up page loads.

Outcome

What I learned

“The technical infrastructure is shared between the two products, while search and citation logic is modelled separately for each body of case law. That split lets both products evolve at the same pace.”