Dragonfly

We’re breaking the memory barrier! 🚀 𝗜𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝗶𝗻𝗴 𝗱𝗮𝘁𝗮 𝘁𝗶𝗲𝗿𝗶𝗻𝗴 𝗶𝗻 𝗗𝗿𝗮𝗴𝗼𝗻𝗳𝗹𝘆: extending RAM with SSDs to handle massive datasets at a fraction of the cost. Read the announcement, deep dive, and benchmarks vs. ElastiCache: https://hubs.la/Q03VQXF40 #DataTiering #ElastiCache #DataInfrastructure

Tired of the endless manual tuning cycle for your databases? 🔄 In this Modern Data Infra Summit talk, 𝗔𝗻𝗱𝘆 𝗣𝗮𝘃𝗹𝗼, Associate Professor of Databaseology at Carnegie Mellon University, explores the evolution of automated database tuning, sharing the groundbreaking work on 𝗮𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗮𝗴𝗲𝗻𝘁𝘀 𝘁𝗵𝗮𝘁 𝗼𝗽𝘁𝗶𝗺𝗶𝘇𝗲 𝗲𝘃𝗲𝗿𝘆 𝗳𝗮𝗰𝗲𝘁 𝗼𝗳 𝘆𝗼𝘂𝗿 𝗱𝗮𝘁𝗮𝗯𝗮𝘀𝗲, delivering peak performance more efficiently and cost-effectively than ever before. Let’s explore how reasoning agents and specialized algorithms are bringing us closer to the reality of truly self-driving databases, revealing why LLMs alone aren’t the full answer (yet!). 🚀 https://hubs.la/Q03VHS5B0 #MDISummit #AI #AITuning #Database #LLMs #MachineLearning #DevOps

Many developers are still surprised to learn that Dragonfly is extremely highly compatible with the Redis protocol and commands. This compatibility, combined with powerful tools like RedisShake, makes the transition remarkably smooth. In this webinar, we’ll show you that even a live cutover is not as daunting as it seems. You’ll learn the practical steps to migrate from Redis (or Valkey) to Dragonfly, unlocking massive performance gains with minimal friction.

Is your in-memory data store ready for the AI revolution? At the Modern Data Infra Summit, Oded Poncz, co-founder and CEO of Dragonfly, discussed 𝗧𝗵𝗲 𝗣𝗮𝘀𝘁, 𝗣𝗿𝗲𝘀𝗲𝗻𝘁, 𝗮𝗻𝗱 𝗙𝘂𝘁𝘂𝗿𝗲 𝗼𝗳 𝗜𝗻-𝗠𝗲𝗺𝗼𝗿𝘆 𝗗𝗮𝘁𝗮 𝗦𝘁𝗼𝗿𝗲𝘀: Memcached solved the disk speed problem. Redis empowered stateless applications. However, as AI demands massive, real-time context, we’ve encountered a new scaling challenge. Discover how projects with modern architectures like Dragonfly are unlocking millions of ops/sec per node and shaping the future of AI-powered applications. Watch the full replay ➡️ https://hubs.la/Q03TyWC60 #MDISummit #AI #ContextEngineering #DataInfrastructure #Performance #Scalability

Dragonfly Search now natively supports geospatial data with the new powerful inverted GEO index. Our latest post breaks down the R-Tree-based implementation and reveals how it delivers dramatically more scalable and performant location-based queries. Read the full deep dive: https://hubs.la/Q03T1xgl0 #Geospatial #Search #DataStructure

Dragonfly is at #KubeCon North America 2025. Come and say hi at booth 2140! 😃

Context building & maintenance, data replay & validation, multi-agent coordination… Sounds like enterprise AI problems? Yes. But they’re also 𝗰𝗹𝗮𝘀𝘀𝗶𝗰 𝗱𝗮𝘁𝗮 𝘀𝘁𝗿𝗲𝗮𝗺𝗶𝗻𝗴 𝗽𝗿𝗼𝗯𝗹𝗲𝗺𝘀. Tyler Akidau, CTO of Redpanda Data, joined our Modern Data Infra (MDI) Summit to unpack how high-performance streaming becomes the backbone for reliable, auditable AI systems in the enterprise. Watch the full session ➡️ https://hubs.la/Q03SkynK0 #MDISummit #DataInfrastructure #RealTimeData #StreamingData

The in-memory data store ecosystem is moving faster than ever. Our new post explores the landscape 1.5 years after the #Valkey fork, the push for performance, features, and innovations, and what it means for building context-rich real-time #AI applications. Read the full analysis: https://hubs.la/Q03S2jr00

Modern context-rich AI-driven workloads require performance and scalability, as the optimal utilization of your servers determines how far an application can go. Valkey’s multi-threaded I/O improves throughput, but its single-threaded data operations remain a CPU bottleneck. For CPU-intensive sorted set commands, I/O threads only gave a ~5% gain. On the other hand, Dragonfly’s true multi-threaded architecture overcomes this limit, achieving 7x higher throughput on the same #AWS 8-core hardware. Read our full analysis: https://hubs.la/Q03RDrxm0 #Valkey #Database #Performance #AI

Is scaling your distributed database a challenging and expensive process? 💸 It doesn’t have to be. Catch this week’s standout talk from the Modern Data Infrastructure (MDI) Summit: “𝗙𝗿𝗼𝗺 𝗖𝗼𝗻𝘀𝗶𝘀𝘁𝗲𝗻𝘁 𝗛𝗮𝘀𝗵𝗶𝗻𝗴 𝘁𝗼 𝗘𝗹𝗮𝘀𝘁𝗶𝗰𝗶𝘁𝘆!” with Felipe Cardeneti Mendes from ScyllaDB. This talk unveils how ScyllaDB’s new tablet-based data distribution and Raft-based metadata management enable dynamic, topology-agnostic scaling, eliminating painful rebalancing and overcommitted nodes for good. A must-watch for any #DevOps or database pro! 🎥 Watch the full replay: https://hubs.la/Q03RpTGZ0 #MDISummit #DataInfrastructure #NoSQL #Performance #Scalability