Today, Shreesha (Shreesha G Bhat) presented our work on Speclog at OSDI 25. Speclog is a new shared log that achieves low end-to-end latencies for real-time applications. (1/8)

I wrote a piece on building 𝗗𝘂𝗿𝗮𝗯𝗹𝗲 𝗔𝗴𝗲𝗻𝘁𝘀 using durable execution, communication, and state in restate. It matches so well, it still sometimes feels too good. The approach works with popular frameworks like AI SDK and OpenAI Developers agent SDK, or even with

We made a Mac app that lets you run a bunch of Claude Codes in parallel. Introducing Conductor!

I'll be presenting this later today - join me for a rundown of how to make agentic workflows easily scalable and resilient.

Many years ago we had a service that experienced a 100% API outage across all hosts in the service, nearly at the same time.

Today is the new semester for CMU Database Group's Intro to Database Systems! We're going harder into material than ever before. Projects are more challenging but you can use LLMs to help. We also have 10min talks each Wed from leading DB companies. Follow on YouTube: 15445.courses.cs.cmu.edu/fall2025

## we know what good UX/reliable software looks like, we just lack good defaults the lack of a “simple, fast, lightweight” orchestrator makes SO much software worse than it could be. we know we should do retries, but it’s extra code and extra infra, so you dont bother. we know

The live timeline still blows my mind every time I see it!

Weekend read: 𝗔𝗴𝗲𝗻𝘁𝘀 𝘀𝗵𝗼𝘂𝗹𝗱 𝗯𝗲 𝗦𝗲𝗿𝘃𝗲𝗿𝗹𝗲𝘀𝘀 (𝗮𝗻𝗱 𝗗𝘂𝗿𝗮𝗯𝗹𝗲) Serverless platforms are amazing for many AI agents, especially if they do a lot of bursty work (chat sessions with long inactivity) or react to events (approvals). Two things are needed

Unikraft Cloud is live. $6M raised. Millisecond cold starts. 100K+ VMs per server. Built for amazing scale and superior unit economics. Read the full release -- link in the first comment 👇

PSA. I’m hiring top-notch software engineers to work with me on restate server/runtime. Rust/C++ experience is highly valued in distributed/storage systems space. Drop me a DM or spread the word ❤️

How do you turn your agent code into something reliable and scalable to run millions of executions? We tried building a cloud coding agent with restate for workflows, Modal for sandboxes, Vercel for compute, and GPT-5 / Claude as the LLM. Think a mini version of cursor

In a few days time everyone will go back to remembering that running your own infra requires highly specialized skills that are a massive distraction for 99% of companies

Building multi-regional systems is hard, evolving existing systems even more so. Restate replicates your application state across regions - not just for DR failover, continuous active replication and automated reconfiguration - with no code changes required.

🔥 Today we’re excited to announce new funding for `grep` (at a $1.3B valuation) to continue building the foundation of agent observability and text search infrastructure. grep began as a humble UNIX utility in 1973. Since then, it’s evolved—through recursive innovation and the

One thing I've learned building and running AWS over the years is that things either need to be limited or billed. In the case of Lambda cold-start time, making it 'free' required awkward limits on time and CPU that added complexity.

PARENTS: please check your kid's candy this halloween - i just found an 18-month enterprise salesforce contract with AI features and a $100k early breakup fee inside this snickers bar

Happy Halloween!