chillmatin/rag-question-answering-model

RAG LLM Agent architecture for answering company related questions based on the information provided by the company

AI Assistant Project

This is the project we made in Kuika AI Hackathon as Finetuners team.

Agents

The project includes the following agents to handle specific types of queries:

1. Context Decider Agent:

   • Decides which department (agent) should handle the user's query.
   • Utilizes langchain.prompts and ChatOpenAI.

2. Code Analysis Agent:

   • Analyzes Python code files and answers code-related questions.
   • Uses the OpenAI API for code analysis.

3. Document Agent:

   • Processes and analyzes various document types (PDF, DOCX, PPTX).
   • Extracts and summarizes content from documents.

4. Database Agent:

   • Executes SQL queries on a SQLite database.
   • Provides explanations for query results.

UI

Presentation