Migrating

Emsallerim

Semantic search engine for Turkish legal precedents and case law.

Built for

  • Lawyers
  • Law firms
  • Law students
  • Legal researchers

What it solves

  • Keyword search over Supreme Court rulings is narrow and incomplete
  • Finding similar cases by hand takes hours
  • Filtering by date, chamber, or decision type is painful
  • No assistant available when you need one

How it works

  1. Supreme Court and Council of State rulings are indexed, each with a semantic embedding.
  2. Your question becomes an embedding, and similar rulings are retrieved.
  3. Filters (date, chamber, decision type) refine the results.
  4. Direct queries via a Telegram bot — results come back as a summary plus relevant passages.

Tech stack

FastAPIQdrantTelegram Bot APILocalAI + nomic-embed-textPostgreSQL

Emsallerim is a semantic search engine that helps lawyers quickly find similar cases before a hearing. Instead of traditional keyword search, it works by legal meaning similarity.

A typical use

You briefly describe the case, e.g., "Can a worker claim severance after a just-cause termination?". Emsallerim returns ten to fifteen strong rulings — each with:

  • Date and chamber
  • Short summary
  • The paragraphs most relevant to your question

Why semantic

Keyword search for "just-cause termination" only returns rulings that contain that exact phrase — you miss equivalent phrasings like "worker's contract ending". Semantic search matches meaning, not words.

Privacy

Your searches stay inside our cluster. No third-party search engine. Client information is never logged.