Got questions? We have the answers.

If you have questions about how this project uses AI for data embedding and search, feel free to contact us.

Contact Us

Our platform collects and summarizes legal documents and information from reliable Danish sources. The goal is to provide brief, precise, and easily understandable summaries of bills, court decisions, and rulings.

We use AI technology to analyze and summarize the content. The system focuses on extracting key facts and context to ensure that the summaries are accurate and correctly represent the original sources.

We primarily use API communication with various databases to extract data. We then use a combination of packages such as html-to-text or pdf-parser, depending on the format of the data we receive from the APIs. For language processing and summarization, we use AI tools like OpenAI's ChatGPT and Anthropic Claude.

AI is used to transform the collected legal content into short, precise summaries. It ensures that the information is easily understandable while preserving essential details from the original documents.

We use Next 14 with React 18, utilizing React Server Components (RSC) and Server-Side Rendering (SSR). For our database needs, we use Supabase PostgreSQL and Pinecone as a vector database.

We collect data from a wide range of official Danish legal sources, including the National Tax Tribunal, the Appeals Board, the Court Database, the Maritime and Commercial High Court, the Labor Court, the Insurance Complaints Board, the Tax Council, the Rent Tribunal, the Appeals Boards Authority, the Public Prosecutor's Knowledge Base, and the EU Court of Justice.

Yes, our platform allows users to closely follow the legislative process by providing updated summaries of bills and amendments. This helps users stay up-to-date with the latest legal developments.

In short, no. You should always verify the actual answers by consulting the original sources.

About the Legal Guide

The project was started in June 2023 as a side project alongside my studies at DTU. In the following semester, I worked in collaboration with DTU Compute to develop this AI Chatbot. The project has since evolved into a search database for various judgments and decisions and legal monitoring.

A special thanks to:

Danske Love for granting me permission to use their legal index to create a useful vector index from which I could query and extract all relevant laws.

Since December 2023, Danske Love has been replaced by the actual legal source Retsinformation.dk.

I am working to make the code open source, so that the RAG, chatbot, and search index will be available to anyone interested.