Project: The Multi RAG application offers diverse system prompt templates suitable for multiple use cases

Project Name

Multi RAG-ing

Description

The Multi RAG application accommodates various system prompt templates tailored for diverse use cases. Leveraging the multilingual capabilities of i18next, it caters to users in English, Spanish, French, and Japanese, enabling prompt interactions in their preferred language.

Key Use Cases:

Shogi Assistant: Supports Shogi learners by answering questions about Shogi and KIF files, providing detailed strategic insights. Enhanced to handle KIF file data ingestion, allowing users to click on citations and review files in a side panel with a KIF player(Kifu-for-JS).
Japanese Adverbs Assistant: Assists with Japanese adverbs, emphasizing reduplicative adverbs (jougo). It provides readings, translations, equivalents in supported languages, and usage examples through engaging sentences.
Contoso Electronics Assistant: Aids company employees with queries about healthcare plans and the employee handbook using sample data from the repo.
Custom Assistant: Offers personalized assistance as a helpful AI assistant across various domains.

Additional Features:

A dynamic question list, shuffled and displayed on the UI, enhances user engagement by presenting a variable set of questions.
A dropdown list to select document categories, showing relevant options based on the chosen prompt template.

The application serves a broad range of users—from Shogi enthusiasts to businesses requiring customer support—through customizable and dynamic interactions.

Technology & Languages

[X] JavaScript
[ ] Java
[ ] .NET
[X] Python
[ ] AI Studio
[X] AI Search
[ ] PostgreSQL
[ ] Cosmos DB
[ ] Azure SQL

Project Repository URL

https://github.com/bnodir/azure-search-openai-demo/tree/RAG_Hack_Project

Deployed Endpoint URL

No response

Project Video

https://youtu.be/weZcEZw-PZ8

Team Members

bnodir

microsoft / RAG_Hack