pierreolivierbonin/Canada-Labour-Research-Assistant

The Canada Labour Research Assistant (CLaRA) is a privacy-first LLM-powered RAG AI assistant proposing Easily Verifiable Direct Quotations (EVDQ) to mitigate hallucinations in answering questions about Canadian labour laws, standards, and regulations. It works entirely offline and locally, guaranteeing the confidentiality of your conversations.

7 stars 0 forks 0 watchers Python

chatbot-application chatbot-framework labour labour-relations lcs-algorithm llm llm-inference llm-serving metadata ollama question-answering quotations rag-chatbot retrieval-augmented-generation sentence-transformers source-referencing streamlit string-matching-algorithms vector-database vllm

View on GitHub

1 Open Issue Need Help Last updated: Jul 5, 2025

Open Issues Need Help

View All on GitHub

Making it possible to disable RAG on demand to use a vanilla model directly 2 months ago

AI Summary: The task is to modify the Canada Labour Research Assistant (CLaRA) application to handle the case where no sources are retrieved (number of sources is 0). Currently, the application throws a TypeError when this happens. The desired behavior is to bypass the Retrieval-Augmented Generation (RAG) process and use the vanilla language model directly if no relevant sources are found, preventing the error.

Complexity: 4/5

good first issue

pierreolivierbonin/Canada-Labour-Research-Assistant

Python

#chatbot-application#chatbot-framework#labour#labour-relations#lcs-algorithm#llm#llm-inference#llm-serving#metadata#ollama#question-answering#quotations#rag-chatbot#retrieval-augmented-generation#sentence-transformers#source-referencing#streamlit#string-matching-algorithms#vector-database#vllm