__STYLES__

Work Wise AI - RAG based AI Assistant on Govt. Service Rules

Work Wise AI - RAG based  AI Assistant on Govt. Service Rules

Demo Video

About this project

Work Wise AI is an innovative Retrieval-Augmented Generation (RAG) based chat application designed to serve as the first line of HR contact for public sector employees. This AI-driven solution is specifically tailored to provide accurate and timely responses to queries related to General Financial Rules (GFR), Fundamental Rules & Supplementary Rules (FR&SR), Central Civil Service rules, and leave policies.

Key Features:

  • Targeted HR Assistance: The application is designed to answer HR-related queries from public sector employees, significantly reducing the time and effort required by HR personnel to handle routine inquiries.
  • Comprehensive Training Data: WorkWise AI is trained on extensive datasets, including GFR, FR&SR, Central Civil Service rules, and leave policies, ensuring that it provides relevant and precise information.
  • RAG Framework: The application leverages a Retrieval-Augmented Generation (RAG) approach, combining the power of retrieval-based models with generative capabilities to deliver comprehensive and contextually accurate responses.

Technical Implementation:

Embeddings with OpenAI: To create meaningful vector representations of the text, OpenAI's embedding model, text-embedding ADA 002 was utilized. These embeddings capture the semantic essence of the documents, making it easier to retrieve relevant information.

  • Indexing with FAISS: FAISS (Facebook AI Similarity Search) was employed to index the embeddings efficiently. This allowed for fast and scalable similarity searches, ensuring that the most relevant documents are retrieved in response to user queries.
  • User Interface with Streamlit: The front-end of WorkWise AI is built using Streamlit, a lightweight and efficient framework that enabled rapid development and deployment of an interactive and user-friendly interface.
  • Retrieval and Generation with OpenAI: Upon receiving a query, the system first retrieves relevant documents using FAISS and then generates a coherent and accurate response using OpenAI's generative model (GPT-4o-mini). This dual approach ensures that users receive well-rounded answers, drawing from the most pertinent rules and policies.

Project Significance:

WorkWise AI represents a significant advancement in automating HR processes within the public sector. By handling routine queries with high accuracy, it not only improves the efficiency of HR departments but also enhances the overall employee experience. This project exemplifies how AI can be leveraged to streamline complex administrative functions, making critical information more accessible and actionable for public sector employees.

Tools and Technologies:

  • Programming Language: Python
  • Embedding and Generation: OpenAI
  • Indexing: FAISS
  • User Interface: Streamlit
Discussion and feedback(0 comments)
2000 characters remaining
Cookie SettingsWe use cookies to enhance your experience, analyze site traffic and deliver personalized content. Read our Privacy Policy.