GKGilad Kleinmaninblog.epsiolabs.com·Aug 10, 2025 · 10 min readWhy is stream processing hard?Before starting Epsio a couple of years ago, I used to spend countless hours learning about creative and innovative ways companies were making data access and querying faster and more efficient. Looking at rising products in that space — whether it’s...00
GKGilad Kleinmaninblog.epsiolabs.com·Aug 5, 2025 · 7 min readThe Bidirectional Stream Processor: Why Pull Beats Push for Crash RecoveryAt Epsio — we’re building a streaming SQL engine, focusing our engineering effort towards ease of use / removing the middleware usually associated with stream processing. As part of that effort — we need to ensure that if and when our stream processo...00
GKGilad Kleinmaninblog.epsiolabs.com·Aug 21, 2024 · 8 min readWhy We Built a Streaming SQL EngineFollowing our previous blog post “How we built a Streaming SQL Engine,” we are delighted to share some of the reasoning and thoughts that led us to build Epsio (as we always say — start with how, only later ask why). Why we built Epsio For most of ou...00
GKGilad Kleinmaninblog.epsiolabs.com·Aug 6, 2024 · 17 min readOut Of Memory ShenanigansAt Epsio Labs, we develop an incremental SQL engine — in brief, an SQL engine that operates on changes instead of reprocessing the whole dataset each time a query runs. I’ll leave explaining why that’s useful to another blog post of ours — Why We Bui...00
