Turso

I have seen many Open Source project leaders complain that the velocity in which code is produced now is deadly: people produce sloppy code trivially and send it upstream, effectively flooding the project's resources and draining the ability to review the code. We at Turso have also been plagued by the same thing. It invites us to think: is Open Source dead? Not for us. We are more Open Source than ever, and doubling down. We see Open Source as a way to give a seat at the table to the people who do the work to push a project forward. The only difference is that what is "work" changed. Writing code is free, and just writing code does not give anyone a claim on our maintainers' time. But many other things do. The beauty of it, is that the new tools democratize access and give a lot more people a chance to participate. Drawing from our experience in the Kernel, and how we think about it from first principles, I wrote some clarifications on our Open Source policy today. Read more: https://lnkd.in/gPDTgk3V

A year ago, we went on a crazy mission: rewrite the most beloved database in the world. We did that because it was clear to us that AI would change the world completely, and we needed a database that had the same file-based nature as SQLite, but could be a lot more powerful and a lot richer. A year later, we improve on SQLite by adding concurrent writes, CDC, encryption, and a long list of other features that make Turso the right choice for all your agentic workloads. At the same time, we turned the Turso Cloud into one of the most reliable database offerings in the market, and added the ability to encrypt each of your databases with your own key. I spent some time today to review the past year, and muse about what the future holds: https://lnkd.in/gCqDhUvW

I wrote some thoughts into the Turso blog today about why I think the Filesystem abstraction will be a key abstraction for LLMs. Read the full thing at: https://lnkd.in/gKWRSDFN A summary: 50 years ago, Unix popularized a concept that seems so obvious to us today, it is hard to grasp how revolutionary it was: "Everything is a file". Unix was built on files. Every single Unix application was designed to consume files, and then spit files. If we look closely, this is one of the first instances of a common interface (as APIs would become later). If there is a common interface, tools can hyper-specialize. And if tools can hyper-specialize, they can become very good at what they do. The many decades of Unix dominance left us with a very wide array of incredibly capable tools that speak the language of files. And late last year, it became clear that "agents" were really just a dumb French guy called "Claude" with access to a Unix shell, commandeering all those tools. The models are trained on those tools. And even for the new tools, it is hard to resist the urge of just following the pattern, and making it file-based at least to some extent. My prediction for 2026 is that filesystems will be back with all the rage. We at Turso developed AgentFS for this exact same purpose. Another tool I will highlight is Just-bash, by Malte Ubl at Vercel: it builds upon the fact that most modern agents are written in Typescript, and gives them access to a subset of the most powerful bash tools an agent would want. We live in wild times, and I am here for it!

Coding agents provide the core architectural pattern that agents in the future will adopt. Amd isolated, safe execution is critical to allow agents to freely explore. In this blog, our friends at LlamaIndex show how to achieve that using LlamaIndex and Turso AgentFS Read more at: https://lnkd.in/gmjhTakv

Move beyond thinking of databases in terms of just capacity and throughput. With AI agents, multitenancy is ending. The age of hyper-tenancy has begun. By Glauber Costa, thanks to ScyllaDB

Today Turso is releasing AgentFS, the best way to handle storage in your agent sandboxes. We have heard from many agent builders, that told us a consistent story: state for agents is hard! Traditional client-server databases won't cut it, because you need very frequent access to data, and want to store assets like files and documents that the agent generates. Traditional filesystems don't work either because they are either not available in some sandboxed environments, or they just can't meet the scale that sandboxed agents require (ever tried to attach an EBS volume to a K8s pod? Imagine if you want agents coming online in milliseconds...) AgentFS provides a rich API for dealing with all of the data an agent needs: Files and Documents, Key-Value for configuration, Audit trails for agent execution state. All of this stored in a single SQLite file, backed by Turso, the next evolution of SQLite. In the future we will see not billions, but trillions of agents throughout the economy. They will come online in milliseconds, and will generate an amount of data that is incomprehensible today. Turso is here to build this future. Learn more about AgentFS: https://lnkd.in/gvG5tAcb