October 2021 Summaries

Citi Bike, New York's popular bike share program with over 20,000 bikes across Manhattan, Brooklyn, Queens, and the Bronx, has a new app that allows users to discover how many bikes are used at different times throughout each day in various neighborhoods. The app uses RedisGraph for lightning-fast queries, enabling users to visualize traffic patterns around the network. Users can drag a geospatial index across different locations to uncover precise figures on Citi Bike usage over time. This application is built using React and around react-mapbox-gl, with an aggregated trip graph constructed using devexpress/dx-react-chart. The offline importer downloads public Citi Bike trip data, unzips each archive, and indexes all of the trips into a journeys graph.

The application aims to simplify task management by providing a clear visualization of project progress and relationships between tasks, allowing managers to plan more efficiently and make quicker decisions in high-pressure situations. It utilizes RedisGraph for storing and querying graph data, enabling efficient transmission of data while projecting visualizations of tasks and their relationships. The application's database stores projects relationally in PostgreSQL, with tasks stored as nodes in Redis Graph and relationships as edges. The app provides a simple interface for users to create, edit, and delete tasks, as well as establish relationships between them. It also includes features such as data seeding, user creation, and task deletion.

The first Feature Store Summit brought together industry thought leaders and practitioners from over 25 organizations focused on feature stores, highlighting the growing importance of this technology in machine learning operations. The event covered key themes such as delivering ML use cases using real-time data with low latency, building vs buying a feature store, collaboration capabilities, robust feature engines, and driving user adoption. Key takeaways included the need for low-latency serving, the hybrid approach using open source solutions, and the importance of instilling trust in the feature store through consistency, killer features, and robust feature engines. The summit also showcased modern feature platforms that provide a solution to close the entire loop between online and offline data, enabling live, fresh, and fast ML features.

Google's search engine dominance creates a barrier for programmers to access niche-specific resources on software development, which are often buried deep within Google's congested library. This leads to the creation of Awesome Search, a search engine that specializes in identifying curated pieces of content from awesome lists, using Redis as its foundation. The app is built with RediSearch, a Redis-based search engine that enables efficient indexing and querying of resources. With Awesome Search, programmers can access valuable coding content that would otherwise be hidden on Google's platform, making it an essential tool for the development community.

The Launchpad App created a language processing machine learning pipeline to weed out confirmation bias in medical literature, utilizing Redis as the data fabric. A knowledge graph was created using RedisGears and RedisGraph to store entities, concepts, and relationships between them. The pipeline uses the Aho-Corasick algorithm to match incoming sentences into pairs of nodes and present sentences as edges in a graph. The system is designed to promote diversity of opinion and prevent confirmation bias in medical professionals' diagnoses. The knowledge graph can be visualized as a graph structure, highlighting each entity's properties along with their relationships. The pipeline uses RedisGears and RedisGraph to process information using RedisGears and stores it in RedisGraph. The system also utilizes the BERT model for summarization and question answering tasks. The Redis Knowledge Graph is designed to create knowledge graphs based on long and detailed queries, allowing users to navigate through medical literature seamlessly without suffering from confirmation bias.

The RedisRaft license has been updated from dual AGPLv3 + RSAL to RSAL only as of October 20, 2021. This change applies specifically to the RedisRaft project and does not affect the core Redis software, which remains under the 3-clause BSD license. The new RSAL license allows for more flexibility in terms of usage and redistribution of source code, permitting users to freely download, modify, and redistribute without copyleft restrictions. This change aims to provide a foundation for enterprise software and services while maintaining the open-source governance of the core Redis project. The company remains committed to supporting the success of the Redis ecosystem through various efforts, including investing in clients and modules under open-source or source-available licenses.

This project utilizes drones equipped with cameras to capture images of crops, which are then analyzed in real-time using Redis and cloud technologies. The analysis is performed by various components such as RedisGears, RediStreams, and RedisAI, which work together to calculate the percentage of different categories in the images captured by the drone. The project also includes a backend system built using microservices that perform data inspection and calculation of sum assured and premium. Additionally, a frontend app is created to provide a user interface for crop insurers to access the real-time data, enabling them to scan, monitor, and assess crop yields from their office.

Enterprises today face a rapidly changing market where the time-to-market keeps shrinking, requiring faster implementation of changes to applications and infrastructure. Organizations must balance stability and innovation while maintaining maximum uptime. A recommended strategy is to follow the "divide and conquer" approach by identifying and prioritizing different system components. Flexible adoption paths enable organizations to upgrade and modify their systems in smaller units of work, based on their goals and requirements. Redis Enterprise v6.2.4 offers flexible upgrade options, allowing users to choose the Redis server version they want to upgrade to according to their upgrade cadence tolerance. The non-disruptive upgrade (NDU) process ensures availability and performance while giving access to the latest Redis Enterprise innovation.

The squad health check system developed by the Launchpad App uses Redis to create an efficient transmission of data, enabling instantaneous feedback, and facilitating a better working environment for businesses. The system allows users to customize questions tailored to their requirements, providing personalized solutions. It features a user-friendly dashboard, automatic session processing, and real-time reporting to Discord. The system's architecture consists of several components, including RedisJSON, RediSearch, RedisGears, and Redis Streams, which work together to provide a seamless and efficient experience for users.

The future of real-time AI is being driven by cloud-based, real-time analytics and AI-driven applications across multiple industries. Redis technology is playing a significant role in this transformation, with its open-source, in-memory data structure store enabling real-time functionality for various applications such as NLP for scientific and medical research. The latest enhancements to the Redis platform, including Active-Active Geo-Distribution and support for feature store functionality in RedisAI, are pushing boundaries in solution value and potentially disrupting major incumbents in the market. As real-time applications become increasingly critical across various sectors, understanding the technical underpinnings that make these advances possible is crucial to shaping tomorrow's application environments.

The gaming industry's growth has highlighted the need for unique and dynamic user experiences, particularly in multiplayer games. To address this challenge, Redis provides a powerful database that supports low-latency gaming use cases, enabling personalized interactions and high-speed reactions. The Launchpad App showcases an online game built using Redis, which uses RedisGears to deploy a sequence of functions based on the chronological order of the application setup. The game's architecture consists of three main components: Redis, NodeJS backend, and Docker compose YAML file. The application uses RedisSearch indexes to enable efficient data transmission between components, allowing for an active-active geo-distributed top-down arcade shooter application to be deployed with exceptional latency speed, making it suitable for interactive gameplay where users from all around the world react and fire missiles at other players.