What does it take to handle millions of messages per second on a single machine? Our CEO, Carl Hörberg, dives into why we built LavinMQ from scratch, why we bet on Crystal over Go or Rust, and why UI matters just as much as raw performance.
LavinMQ
@LavinMQ
Lightning-fast message broker designed for high-volume messaging.
Joined May 2022
- Operators don't open infra dashboards casually, they do it because a queue is spiking. The LavinMQ management UI is designed for high-pressure troubleshooting. Here is how we prioritize radical visibility and keep the UI lightning-fast even at massive scale.
- TrueTale built an AI tool for authors, but massive data ingestion was overloading their servers and wasting expensive AI tokens. By moving to an async setup with LavinMQ, they cut their server workload by 2/3 and tamed traffic spikes for good. lavinmq.com/blog/truetale-…
- Building a RAG-powered chatbot? In part 2 of our LangChain and LavinMQ series, we move from architecture to implementation. Learn how to decouple your data ingestion from LLM processing to build a resilient AI system that doesn't buckle under high traffic.
- One of the biggest hurdles with AI is ensuring it stays grounded in your specific data. In this tutorial, we show you how to build a RAG-based assistant using LavinMQ and LangChain so it uses your actual documentation as its absolute source of truth.
