LAW-IO-BUDDY is a legal assistance tool which provides 3 main features followed by several customization options to enhance legal research. The 3 features are: 1) Web Extension to Interact/query websites - This tool is an interactive Web Extension for google chrome which can scrape all the text from the website to interact with it. Users can ask questions related to the webpage to get concise answers through ChatGPT. The scraped data is first stored in Vectara and using a combination of Semantic Search and ChatGPT, the user can get answers to all their queries in the webpage. 2) Question & Answer with Documents - Using an interactive webpage developed on streamlit, the user can upload their files to query upon them. But this gets even better when the user selects the customization options, the user can query with their documents while combining the information from their documents with legal documents of multiple countries to get concise information to verify if their document related query is resolvable by the Legal Database of the country. Moreover, they can select the length of the output as well find the source of the answer in the document. 3) Legal Chatbot - The final piece to the ultimate legal research assistant, the legal chatbot can answer any law/legal questions from its database of various constitutional and legal documents stored in its database. It also has multiple customization options, which include "Search the Web", which is a novel feature which can combine web related searches to legal documents to provide accurate answers to user queries. Moreover the user can also select the length of answer while checking its source. Overall, LAW-IO-BUDDY is the ultimate legal research assistant which can boost the speeds of users during research whilst also being a beginner friendly platform to accommodate users who are unaware about the intricate aspects of certain laws. The 2nd and 3rd features can also run as standalone webpages.
Category tags:"excellent work"
Walaa Nasr Elghitany
Data scientist and doctor
"- Would have been better to set out the business context for the use case first, before jumping into the technical details, features, and demo. - Good approach of combining different UI modes (chatbot, Q&A), different tools (browser extension, web app, vectara, chatgpt), and query the docs as well as the web. - Would have been good to hear an explanation of what the reference numbers are for in the summary from vectara."
Justin Hayes
Head of Field Engineering