Verba: Simplifying Data Analysis with AI
Verba is an open-source project that aims to provide a simplified and user-friendly interface for RAG apps. It allows users to dive into data and have meaningful conversations quickly.
Why Verba is More Than Just a Tool
Verba is not just a tool for querying and manipulating data; it is a companion that facilitates paperwork, data comparison, and analysis. By leveraging Weaviate and Large Language Models (LLMs), Verba enables users to perform these tasks effortlessly.
The Power of Weaviate and LLMs
Verba is built on Weaviate’s cutting-edge Generative Search engine. Whenever a search is performed, Verba automatically extracts relevant background information from documents using the power of LLMs. This allows Verba to provide comprehensive and context-aware solutions. The straightforward layout of Verba makes it easy to retrieve and explore this information. Additionally, Verba supports various file formats for data import, including .txt and .md, and automatically processes the data for better search and retrieval.
Optimizing Verba with Weaviate’s Features
Verba takes advantage of Weaviate’s create module and hybrid search options. These advanced search methods scan through papers to find important context pieces, which are then used by LLMs to generate in-depth responses to inquiries. Furthermore, Verba enhances search speed by embedding both the generated results and queries in Weaviate’s Semantic Cache. This allows Verba to quickly find similar answers in the cache before responding to a new question.
Getting Started with Verba
To use Verba, you need an OpenAI API key for data input and querying capabilities. The API key can be added to the system environment variables or included in a .env file when cloning the Verba project. Verba offers flexibility in connecting to Weaviate instances based on your specific use case. If the VERBA_URL and VERBA_API_KEY environment variables are not present, Verba will default to using Weaviate Embedded, which simplifies the process of launching the Weaviate database for prototyping and testing.
Before importing data into Verba, keep in mind that it may incur costs based on your OpenAI access key configuration. The Verba project exclusively uses OpenAI models, and the API key will be charged for the usage of these models. The primary cost drivers are data embedding and answer generation.
To experience the power of Verba, visit https://verba.weaviate.io/.
The Three Components of Verba
- Host your Weaviate database on Weaviate Cloud Service (WCS) or your own server.
- Utilize the FastAPI Endpoint to mediate between the Large Language Model provider and the Weaviate data store.
- Explore and manipulate data using the React Frontend, a dynamic user interface delivered via FastAPI.
For more detailed information and to try out Verba, visit the GitHub repository and give it a try. Please note that all credit for this research goes to the dedicated researchers behind the Verba project. Join our ML SubReddit, Facebook Community, Discord Channel, and Email Newsletter to stay updated on the latest AI research news and projects. If you enjoy our work, you’ll love our newsletter!
Dhanshree Shenwai is a Computer Science Engineer with experience in FinTech companies. She has a keen interest in the applications of AI and is passionate about exploring new technologies and advancements that make life easier for everyone.