Agents re-read datasets— a lot. The "data leaving storage" is real. @crewai.bsky.social + Tigris fixes that. • Unlimited reads, no cost penalty • Durable workspace across agents + runs • Snapshots for reproducible state Storage built for how agents actually behave. Guide 👇

AI agents break most infra assumptions. They pause. Resume. Fork. Snapshot. Move regions. That’s not stateless microservices. We’ve been working with @agentuity.com as they build a new cloud for AI agents — and their storage model looks very different

Running MinIO yourself avoids big cloud egress fees—but with MinIO retiring its open source project, many teams are rethinking object storage. Tigris checks many boxes for compliance: • We sign BAAs (healthcare) • Region restrictions for data residency • SOC 2 Type II compliant • Fine-grained IAM

We open-sourced our Agent Skills Library — reusable skills that teach AI agents how to use the Tigris SDK, convert images to webp, write conventional commits, and more. Install them all: npx skills add github.com/tigrisdata/s... Read more about it 👇

Neoclouds are challenging hyperscalers with a distributed computing architecture-modeled storage layer that supports computing across any cloud. Thanks to @tigrisdata.com

Building a hosted MCP server? Auth shouldn't be the hard part. We open-sourced our MCP OIDC provider, a minimal, vendor-neutral OIDC layer that works with real MCP clients (Claude, Cursor, etc.) and any IdP. Blog + code 👇

If you're a developer, you live in the command line. Tigris' new CLI lets you supercharge your shell with bucket forking, object management, and more all from one easy to use program. It's just an npm install away: npm install -g @tigrisdata/cli

Bom dia all! Xe here and I'm in Portugal this month! Are you ready to Go? Tigris is!

Due to an upstream DNS provider issue, t3.storage.dev will return NXDOMAIN errors. Please migrate your workloads to use t3.storageapi.dev instead. We are actively working on the situation. Follow this thread for more updates: status.tigris.dev/incidents/01...

You've asked for it and now you've got it: Tigris has a Discord server! www.tigrisdata.com/blog/discord... Enjoy this mildly hinged Xe post about how they square peg -> round hole'd the backlog of question/answer data into Discord. Stay safe y'all!

👀 skills.sh/tigrisdata/s...

👀 🔜 discord.gg/puVmEmADzE Details to come soon.

Fun fact: object storage (AKA: cloud based unlimited FTP server) is actually a key-value store under the hood! There's nothing stopping you from using it as one! Today's blogpost covers the gorey details about how and why you'd want to do that.

More teams are moving out of Big Cloud to neoclouds better built for AI workloads. How does the growth of neoclouds change the storage layer? Our CEO talked with @thenewstack.io about the new cloud and how distributed computing is taking over and what that means for storage. 🔗👇

What a year! From $25M funding to Git-like bucket forks, our 2025 was all about powering modern AI workflows with fast, flexible object storage. Catch the highlights! 👇

Chunking is a primary control knob for RAG quality. Compare chunking strategies using a reproducible pipeline: • @qdrant.bsky.social for per-experiment indexing and retrieval diffs • Tigris (Forked Buckets) for dataset versioning • CrewAI to orchestrate ingest → chunk → embed → evaluate Code 👇

Tigris lets you fork your buckets like you fork your code. How did we do it? We made our data plane append-only. To do this you have to make your log (of objects inserted) into the database.

2025 sure was something eh? Next year give yourself the ultimate new year's resolution: switching to object storage with unhateable IAM. Take that, infrastructure Sisyphus! Read more on the blog: www.tigrisdata.com/blog/iam-pol...

Availability is more important than ever. You don't have to lock into a big cloud to get 99.99+ % availability-- Tigris makes the same guarantees, with public data to back it up. We made our availability metrics public: bytes up bytes down, no funny business. Check out the live dashboard 👇

Tigris delivers 4× higher throughput than S3 and 20× higher than R2 for small object workloads — 5.3× / 86.6× faster on reads. 🔥 Running training / inference on CoreWeave, Together AI, Lambda Labs, etc? You don’t need to stay locked into one big cloud to get performance. Run the benchmarks👇

@pixeltablehq makes multimodal data feel like Postgres: declarative tables with raw data, embeddings, captions & model outputs, all queryable with SQL-like syntax. Pair it with @tigrisdata and all your data is hot. No caching. Fast queries everywhere. Native search & retrieval in 7 lines. Code 👇

We just open-sourced our MCP OIDC provider 🚀 If you’re building a hosted MCP server, auth shouldn’t be the hard part, especially with how quickly MCP changes. This gives you a simple, vendor-neutral OIDC layer that works with real MCP clients (Claude, Cursor, etc.) and any IdP. Blog + code 👇

Tired of miles of boilerplate code for simple storage operations? We got you. The Tigris Storage SDK makes simple things simple: put, get, list, delete - because using storage in your app should not be hard. Link in reply

Tired of public registry limits, egress fees, or compliance headaches? You can now spin up a private Docker registry on @northflank backed by Tigris in 1 click. Type in your keys, deploy, done Easy scaling, multi-region, zero hassle. Full guide in the replies.

Whole-bucket snapshots on Tigris = git tags for your data Time-travel your object storage, roll back mistakes, and clone massive datasets for safe experiments. Great alongside existing data-versioning tools, or as a lightweight starting point. 🔗 Full post in reply

MCP servers need to authenticate clients so you can safely provide access to protected resources. The industry is settling around using OAuth to do this, but how do you set it up? Here's what we learned making our MCP server work and how you can get started with your own OAuth flow.

If you want to use Tigris with your AI agent of choice, you can now do it with our hosted MCP server! This lets you get all the Tigris-y goodness you know and love from any AI agent you want. No install needed. This also means you can use Tigris from ChatGPT on the web!

Aaron Francis captured it perfectly on Database School: with Tigris, ‘the region is just Earth.’ No CDNs, no cross-region replication, no complex DR— Reliability is the default on Tigris. Link to the full episode below 👇

Bottleneck alert: traditional Kafka struggles at global scale. Enter AutoMQ + Tigris: stateless brokers, zero local disk, separate compute/storage, and no egress fees. Read more on the blog: www.tigrisdata.com/blog/scaling...

Our partner LanceDB spotted us at Kubecon! We've been building cool things with Lance: 💻 Train directly off object storage while using your full GPU with Lance look-ahead caching and Tigris' low latency storage ➡️ Store vector embeddings alongside your data with Tigris backing Lance Links in reply

We took a "snapshot" of our bucket here at Kubecon Tigris has full bucket snapshots with point in time recovery. No messing with per-object versions. We'll keep all your objects safe with 11 9's of durability... even those raffle tickets.

Classical clouds are obsolete. The future belongs to specialized clouds built for AI and high-performance workloads — but they still need reliable, S3-grade storage. That’s where Tigris comes in. 📷 Great chat between @ovaistariq.net and Martin Casado on the a16z podcast, link in reply

If you're at Kubecon this week, we're at Kubecon this week! Check us out and tell us your favorite thing about object storage!

Thanks to Martin Casado for having @ovaistariq.net on the a16z podcast to talk about the future of infrastructure for AI, and how Tigris is building the storage layer for AI workloads— from the metal up. Learn why AI companies keep choosing Tigris for their storage! 🔗 in reply

Big thanks to Aaron Francis for having Tigris CEO @ovaistariq.net on Database School! As Aaron said, “It’s a bold undertaking to go against S3.” We couldn’t agree more — and we’re here for the challenge. 🎧 Catch the full episode to see how Tigris is tackling vendor lock-in! 🔗 Link in reply

Tigris Data is proud to sponsor #RaySummit, where the next generation of AI is being built! Come meet the Tigris team and let’s talk object storage for agents, pipelines, and experimentation— the storage layer that scales with you! 📍 Marquis Marriott in SF today & tomorrow!

Tigris has bucket forking, but how did we do it? We made objects immutable and sorted those object versions by timestamp. Backwards. We downloaded our entire database engine, we're using our entire database engine! Read more on the blog: www.tigrisdata.com/blog/bucket-...

Bucket forking allows you to fork data as easily as code in S3-compatible object storage, without copies, delays or extra costs. Thanks to @tigrisdata.com

Check out the write up @thenewstack.io did on how Tigris Data's bucket forking allows you to fork data as easily as code without copies, delays or extra costs. Read the article on The New Stack now! 🔗 Link in reply

Experiment without risk. Have a big production dataset? Don’t copy—fork it. Make alternate timelines for each model or transformation with zero copying. We tried cleaning, transforms, and inference—no impact on source. 🔗 Code on the blog, link in reply.

“The log is the database; everything else is cache or materialized views.” That idea inspired how we built Tigris– an object store where the append-only log is the system itself. Let’s unpack what that means for the future of storage.

We built bucket forking so agents can safely use tools and run experiments without corrupting shared environments, especially when running in parallel. Learn how every agent gets an instant, isolated sandbox to test, iterate, and scale without conflicts. www.tigrisdata.com/blog/build-a...

The keynotes are packed, the coffee is hot, and the Tigris Data team is ready and rockin' at Booth B5 at #PyTorchCon in SF! Come meet the team today & tomorrow and learn all about our new, first-of-its-kind feature, bucket forking! Fork data like you fork code! #AI #ML #PyTorch

We built `git fork` for your data. Instant, copy-on-write clones of your datasets. ZFS + Git = Tigris Storage Thread: 1/10

Come visit Tigris at Booth B5 at #PyTorchCon! Next Tues/Weds, Oct 22-23 in Downtown San Francisco! Tigris + PyTorch opens a world of opportunities like streaming massive datasets during training bit.ly/3Jh5JBa Come see what else is new in #OpenSource #AI & #ML and join the conversation!

Training data? In your object storage? It's more likely than you think thanks to Tigris! Read more on the blog: www.tigrisdata.com/blog/pytorch...

Come meet Tigris at #PyTorchCon, Oct 22-23 in SF 😎 This is THE event to gain early access to the latest #OpenSource AI tools & contribute to pivotal discussions that will shape the future of #AI projects. Check out the schedule: bit.ly/4cjkIE6. Register NOW: bit.ly/3R2uXUy.

Will we see you at #PyTorchCon, Oct 22-23 in San Francisco? Here are the five talks Tigris is most looking forward to, each one showcasing performance optimizations for AI workloads. See all these and more, and come chat with Tigris about how we're building the storage layer for AI!

Tigris Data is a Startup sponsor for #PyTorchCon, Oct 22-23 in San Francisco! The schedule is STACKED with engaging presentations & panel discussions as well as opportunities to network with the #PyTorch community. Explore the schedule: bit.ly/4cjkIE6. Register today! bit.ly/3xCZrWm

Big news! @tigrisdata.com raised $25M Series A led by Natalie Vais of Spark Capital w/ Martin Casado of a16z + Basis Set. Tigris is a faster, freer, AI-native storage cloud– running on our own hardware. Already powering fal.ai, krea.ai & Hedra. Let’s build the open cloud for AI. Read more on TC