November 2023 Summaries

The text discusses the importance of database uptime and how Weaviate, a production-ready database, addresses this issue through replication. Replication ensures redundancy and prevents user-level downtime during node-level maintenance or updates. Enabling replication in Weaviate is easy and can be done by modifying the configuration file to enable multi-node setup and setting the appropriate replication factor. The benefits of replication include improved availability, reduced critical downtime, and increased redundancy. However, there are trade-offs such as increased hardware requirements and longer start-up times. Weaviate recommends a replication factor of 3 for optimal performance and availability.

Retrieval Augmented Generation (RAG) is a popular application of Large Language Models (LLMs) and Vector Databases that involves augmenting inputs to an LLM with context retrieved from a vector database like Weaviate. RAG applications are commonly used for chatbots and question-answering systems. Evaluating the performance of RAG is crucial, and it involves three components: indexing, retrieval, and generation. Recent advances in using LLMs to evaluate RAG systems have accelerated their development. This article presents an overview of RAG metrics, tunable knobs, experiment tracking, and the transition from RAG to Agent Evaluation.

UK-based startup Moonsift is utilizing AI technology, specifically machine learning models and Weaviate's vector database, to enhance product discovery for online shoppers. The company offers an ecommerce browser extension that enables users to curate shoppable boards with products from various internet retailers. Co-founders David Wood and Alex Reed aim to improve product discoverability by leveraging AI capabilities. Moonsift has collected data on over 60 million products, 250 million interactions, and 40 thousand retailers across the internet, which it uses to train machine learning models for better product discovery. The company is now preparing to launch its AI Copilot feature, which can understand user intent and provide intelligent search results. Moonsift plans to continue optimizing its system's cost, performance, and scale while exploring new features like Product Quantization (PQ) and multi-tenancy for personalized vector search.