Building an AI Study Assistant with Unbody & NextJS
From PDFs to videos, from semantic search to rerankers—learn how to build an AI-native assistant powered by RAG without even knowing what RAG is, all in 2 hours.
Join us for a hands-on workshop where you’ll learn how to build a personalized AI study assistant using Unbody. In this session, we will focus on creating a truly intelligent assistant that leverages the power of large language models (LLMs) to provide a human-like interaction experience. This AI assistant will help students with their day-to-day tasks such as managing assignments, conducting research, and summarizing content. Such a product can be used in various domains including education, professional training, and beyond.
By the end of this workshop, you will have built a fully functional AI-native assistant app from the ground up, incorporating a Retrieval-Augmented Generation (RAG) system to provide advanced search and generative capabilities.
Problem:
Manual Tasks: Students spend a lot of time on manual tasks such as copying/pasting, organizing materials, and extracting content.
Time Consumption: Reviewing and analyzing study materials is time-consuming even with tools like ChatGPT, which lack specific knowledge of the user's materials.
Limited Integration: Existing tools do not seamlessly integrate with various data sources and file formats, limiting their effectiveness.
Solution:
AI Study Assistant: An intelligent assistant that automates transcription, content summarization, and integrates seamlessly with data sources like Google Drive and Google Calendar. It supports various file formats, making it a versatile tool for students and professionals alike.
What You Will Learn by the End of the Workshop:
In this workshop, you will learn how to use Unbody to build an AI-native assistant app with an RAG system without having to know what RAG is. Here is a concise list of the topics we will cover:
Basic Retrieval-Augmented Generation (RAG)
Semantic search (Retrieval); Implement advanced semantic search to find and retrieve relevant content from various data sources, including PDFs, images, audio, video, and Google Docs.
Generative and Prompt-Based Search (Generative and Augmentation); Utilize generative AI to create meaningful responses based on retrieved content and enhance interactions by using prompt-based techniques.
Advanced Features
Advanced filtering; Learn to apply advanced filtering techniques to refine search results based on metadata, relevance, and user preferences.
Using rerankers to improve search; Implement reranking algorithms to reorder search results, improving the accuracy and relevance of the retrieved content.
Integrating various data sources; Seamlessly integrate and manage data from multiple sources such as Google Drive, Google Calendar, and other file formats, ensuring up-to-date and comprehensive knowledge base.
Supporting Multimodal Inputs; Add support for text, image, and audio inputs in your application to enable rich and versatile interactions with the AI assistant.