See how Husky enables interactive querying across 100 trillion events daily by combining caching, smart indexing, and query pruning.

Interactive features without events

R2 SQL provides a built-in, serverless way to run ad-hoc analytic queries against your R2 Data Catalog. This post dives deep under the Iceberg into how we built this distributed engine, from its metadata-driven planner to its parallel execution model.

2 Months of Banging My Head Against Send/Sync, Explained in One Coffee Break

The rapid development of large language models (LLMs) has led to the widespread deployment of LLM agents across diverse industries, including customer service, content generation, data analysis, and even healthcare. However, as more LLM agents are deployed, a major issue has emerged: there is no standard way for these agents to communicate with external tools or data sources. This lack of standardized protocols makes it difficult for agents to work together or scale effectively, and it limits their ability to tackle complex, real-world tasks. A unified communication protocol for LLM agents could change this. It would allow agents and tools to interact more smoothly, encourage collaboration, and triggering the formation of collective intelligence. In this paper, we provide the first comprehensive analysis of existing agent protocols, proposing a systematic two-dimensional classification that differentiates context-oriented versus inter-agent protocols and general-purpose versus domain-specific protocols. Additionally, we conduct a comparative performance analysis of these protocols across key dimensions such as security, scalability, and latency. Finally, we explore the future landscape of agent protocols by identifying critical research directions and characteristics necessary for next-generation protocols. These characteristics include adaptability, privacy preservation, and group-based interaction, as well as trends toward layered architectures and collective intelligence infrastructures. We expect this work to serve as a practical reference for both researchers and engineers seeking to design, evaluate, or integrate robust communication infrastructures for intelligent agents.

This Berkeley systems group paper opens with the thesis that LLM agents will soon dominate data system workloads. These agents, acting on b...

We launched a new product called WarpStream Tableflow that is the easiest, cheapest, and most flexible way to convert Kafka topic data into Iceberg tables with low latency, and keep them compacted.

Hummock is a cloud-native LSM storage engine developed by us. In this article, we will introduce some optimizations that Hummock has made for streaming.

<img src="https://media.licdn.com/dms/image/v2/D4D08AQFRkRFMl4JfiA/croft-frontend-shrinkToFit767/B4DZhTZuRuHsAQ-/0/1753745911803?e=2147483647&v=beta&t=ZSK-8rwkoizGzDG3CIDTLfiaj9_mOqlL91DuqOSg43M" /><br/>

Open-source search and retrieval database for AI applications. - chroma-core/chroma

wal3, Chroma's write-ahead log, combining a 30-year-old lock-free algorithm with S3's newest conditional writes feature.

This is a data system internals blog post. So if you enjoyed my table formats internals blog posts , or writing on Apache Kafka internals or Apache BookKeeper internals , you might enjoy this one. But beware, it’s long and detailed. Also note that I work for Confluent, which also runs Apac

Embedding sizes and architectures have changed remarkably over the past 5 years

Why do we use caches at all? Can databases fully replace them?

A data-driven comparison of MCP and CLI approaches for coding agent terminal control

How we validate strong consistency

On GPT-5 being underwhelming, scaling laws and where we can get reasonable progress from. Also, should AGI still be the goal?

Discover how we reengineered our metrics storage engine for massive scale with Rust, a shard-per-core model, and real-time performance.

Open Source Kafka users get native Apache Iceberg™ topics that collapse ETL and reduce costs - one write to the data lake, many reads - zero copies.

A deep dive into evals, goldens, and hill climbing for improving coding-capable AI models.

<img src="https://miro.medium.com/v2/resize:fit:1050/0*1hSoKkktmPABPkTj" /><br/>

How to make sense of all the data generated by AI Agents.