https://github.com/aleksbrsc/legal-lens
A tool for processing physical or digital multilingual documents into organized data using AI (frontend demo hosted on Pages)
https://github.com/aleksbrsc/legal-lens
eden-ai figma-design gpt-3-5-turbo html-css-javascript tesseract-js
Last synced: 2 months ago
JSON representation
A tool for processing physical or digital multilingual documents into organized data using AI (frontend demo hosted on Pages)
- Host: GitHub
- URL: https://github.com/aleksbrsc/legal-lens
- Owner: aleksbrsc
- Created: 2024-01-20T06:31:19.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-01-28T08:49:59.000Z (over 1 year ago)
- Last Synced: 2025-02-01T20:31:27.581Z (4 months ago)
- Topics: eden-ai, figma-design, gpt-3-5-turbo, html-css-javascript, tesseract-js
- Language: JavaScript
- Homepage: https://aleksbrsc.github.io/Legal-Lens/
- Size: 60 MB
- Stars: 3
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Legal Lens
A tool for processing physical/digital legal documents of many languages into key-value organized data.
### Inspiration
Applying for many American government documents is far too time-consuming and inconveniences those whose native language is not English. Immigrants and newcomers to English countries needed a solution to save their time and energy in this process.
### What It Does
Legal Lens was built exactly to solve this problem. It is a web app that performs image-to-text transcriptions for legal documents in over 100+ languages, translates them into English, and then organizes the textual data in a straightforward, key-value format. The user can then paste this simplified data into their applications, use it in their auto-fills, organize and save their personal information digitally, etc.
You can select the document's language, drag an image of it onto the screen (you may even copy-paste words into the textbox / use voice dictation in your native language), and click a button to convert the translated English text into a parameterized format similar to legal applications.
### How We Built It
This project was made in 2 days using plain HTML/CSS/JS. It uses Tesseract.js for image-to-text transcription, EdenAI for translation, and OpenAI's GPT-3.5 Turbo API for organizing and processing text to send back to the user.
### What's Next for Legal Lens
If we were to revisit this:
- speech-to-text + language recognition
- multilingual website
- web extension/browser plugin for autofill
- more focus on mobile, option to make it work on camera