Case study · Product · Legal SaaS
A multi-tenant SaaS where every legal deadline explains how it was calculated, every firm’s data is isolated by the database itself, and AI suggests while the lawyer decides.
How we made it
A multi-tenant SaaS where every legal deadline explains how it was calculated, every firm’s data is isolated by the database itself, and AI suggests while the lawyer decides.
Case study · Product · Legal SaaS
We sat with the people doing the work and wrote down how it runs today, and where it hurts.
The problem
Law firms in the region track hearings, appeal windows and powers of attorney across spreadsheets, calendars and memory. Generic practice software is built in English first, struggles with Arabic names and Hijri dates, and asks lawyers to trust dates it cannot explain.
We set three rules before writing code: no deadline without a written rule and its legal source; no generated deadline counts until a lawyer confirms it; and no firm can ever see another firm’s data, even through a coding mistake.
What we built
Pick one to see it
Each deadline stores the rule, its source and every step of the calculation, and starts as “pending confirmation”.
Application scopes plus PostgreSQL row-level security, forced on every protected table and verified after each deploy.
Names normalised and compared with trigram similarity, tuned so near-matches are shown, not hidden.
Money in exact integers, an append-only trust ledger with a database constraint, gap-free invoice numbers.
Consent per firm, a monthly ceiling, every call logged; extraction saved as pending; contract review against a checklist.
A separate login and panel; documents and messages are hidden unless a lawyer releases them.
Tell us how your team works today. We will tell you honestly what we would build, or whether an existing product is enough.
Ask the studio
Hi! Ask me what we build, how we work, or about any project on this site.
Answers come only from this website. For cost and timing, send a short brief.