How to Search Supreme Court Case Law with Semantic AI

The Problem with Searching Case Law in Spain

Every lawyer knows the frustration of searching CENDOJ. The database of Spain's General Council of the Judiciary is essential, but its keyword search engine has significant limitations, it only finds judgments containing exactly the words you typed.

Semantic Search: Search by Concept, Not Keywords

AI semantic search works fundamentally differently. Instead of matching text, it understands the meaning of your query and finds conceptually related judgments.

Practical example: A semantic search for "judgments annulling apartment sales due to hidden defects when the seller was a real estate agent" finds judgments using "rescission," "defect in consent," "vendor fraud," etc., even if they don't contain your exact words.

How Lexiel's Search Works

Lexiel's case law system uses vector embeddings over a corpus of Supreme Court, Provincial Courts, and TSJ judgments. Your query is converted to a semantic vector, matched against the vector database, and the AI explains the relevance of each result to your specific case, while verifying citations against CENDOJ to confirm they're real.

Anti-Hallucination Verification

A crucial difference from using ChatGPT for case law research: Lexiel verifies that every cited judgment actually exists in CENDOJ. If it cannot verify it, it tells you clearly. You never risk presenting invented case law.