The file system that connects AI to its data. We power the agents, platforms, and models you love.
San Francisco, CA
- Tweets 83
- Following 15
- Followers 2,301
- Likes 70
Grow the disk of your @AmpCode orbs to infinity by leveraging the new integration with Archil. Archil makes it simple to build orbs that can grow to terabytes of data, but still resume quickly.
Check out the docs here: ampcode.com/docs/orbs/custom…
1
23
6,088
Today, we're excited to introduce Archil Persistent Sandboxes, making it simple to build full-stack agentic applications without having to leave the Archil cloud.
They are the ideal way to run long-running code, like agent loops, alongside context stored in Archil disks.
3
3
36
16,004
Read more in our launch announcement here: archil.com/post/introducing-…
Persistent Sandboxes are available today to all Archil customers in AWS regions, and can be deployed into BYOC environments (along with Archil disks) in about an hour.
15
20,285
celld and Durable Objects that have a full file system, so you can run anything inside of them.
I love celld/DOs, but it’s always annoyed me that state is restricted (somewhat arbitrarily) to SQLite. I’ve been messing around with @archildata to give a full, distributed filesystem to every cell (shimming node:fs). A fun side effect: throughput is ~15x higher than SQLite.
2
9
2,579
We're thrilled to participate in @computesdk's storage benchmarks, and we're delighted to be the highest-performance storage option on the market right now.
Because this tests Archil through our object-storage compatibility layer, file system users will see even higher perf!
4
29
18,323
Archil retweeted
Since we started @archildata, we knew that @huggingface was an important repository of data for AI research and training workloads. Even though HF has great PyTorch loaders, it's still annoying to *work with* the data inside of HF, and research teams need a live file storage for iteration speed.
Today, we're introducing support for @archildata disks to natively synchronize data to HuggingFace buckets (with HuggingFace data sets and models close behind), which makes it easy to get local-like speed to your HF data.
Try it today on the Archil console!
2
2
27
2,268
Archil retweeted
We're excited for @archildata to be hosting a YC Startup School afterparty at our office in Union Square. Come to talk about file systems, eat sushi, and get $10K in credits to build agents on top of Archil.
RSVP via the link on your Startup School attendance confirmation, and see you soon!
3
16
4,516
Agents don't work with infrastructure, they work with tools and context. Watch @s4chinraja demo how Archil's agent framework integrations make it easy to build the context for your agent, then hand agents the tools to work with that context -- with just 1 line of code.
2
10
3,529
See more in our documentation here: docs.archil.com/integrations…
2
511
Archil retweeted
idk if the world is ready for a fast, cheap, serverless filesystem like Archil
very disruptive stuff
4
28
3,358
Agents need structured data too. Developers are using SQLite to store agent history, state, and memories, alongside Markdown.
Today, we're announcing serverless SQLite on Archil. Create unlimited databases on your disk, and use our SDK against them -- from multiple clients.
1
19
10,839
See more information in our documentation here: docs.archil.com/compute/sqli…
3
548
Agents work best with context that’s in one place: a file system. Today, we're announcing Archil agent tools. With a single line of TypeScript, you can hand agents the tools to use Archil's infinite disks, including reading, writing, and running code on files.
2
20
6,030
See more information in our announcement post here: archil.com/post/context-that…
1
414
Archil retweeted
Storage SDK now has support for @archildata.
With native snapshotting and forking coming soon!
1
1
8
7,149
Today, we're excited to announce the first-party Mastra integration for the Archil file system.
This integration makes it easier than ever to select Archil as the storage for your agents, giving you portability across sandbox providers, branching, and the ability to run serverless Python/Node/Bash on your file system directly.
2
5
26
11,219
See more information on the @mastra documentation here: mastra.ai/reference/workspac…
2
736
Today, we're excited to announce that Archil is available as a native adapter in the files-sdk.
This simplifies adding Archil to applications which already use S3 storage, but also need to branch buckets or run bash natively over their data.
1
2
27
4,207
Thanks @haydenbleasel for your help getting us included!
See more information in the docs, available here: files-sdk.dev/adapters/archi…
2
567
Today, we're excited to announce the atomic AppendObject API for all Archil disks. This makes it simple to build agents that are using logs as their fundamental storage medium, Kafka-like streaming workloads, or use Archil as the data source for application logging.
1
2
19
14,043
The Archil AppendObject API is available today for all disks in all cloud regions. See more in our documentation here: docs.archil.com/protocols/s3…
5
773