Object storage: where the files actually sit
Not a disk and not a database. You put a file in under a name, you get it back by that name, and the thing underneath grows as far as you need without you managing it.
What is object storage?
A file system gives you folders and lets you change part of a file in place. Object storage gives you a flat space of keys and whole objects: you write the object or you replace it. In exchange you get durability measured in many nines, no capacity to plan, and a URL for everything.
In an AI system it ends up being the floor everything stands on. The training data, the source video, the documents fed to a model, the generated results, the logs. Almost every pipeline step reads from it and writes back to it, which is why it is also the first place a data-residency conversation lands.
Four ways to run it
The S3 API became the common language here, so all four speak it and most tools can be pointed at any of them by changing an endpoint.
Amazon S3[1]
The one the API is named after. Storage classes let a file get cheaper as it gets older, lifecycle rules move it there automatically, and a signed URL hands out one object for a limited time.
Official documentation →Azure Blob Storage[2]
The same shape on Azure, with its own tiers and its own access model built around the Entra identity your organisation already has.
Official documentation →Cloud Storage[3]
Buckets with a choice of how widely a bucket is spread, from one zone to several regions, which is the dial you turn for durability against cost.
Official documentation →MinIO[4]
S3-compatible storage on your own hardware. Because the API matches, the same code can write to a bucket in your building and a bucket in a cloud, which is what makes a hybrid design practical rather than theoretical.
Project documentation →Where I used it
The video pipeline
Source footage goes in, every intermediate file passes through, and the finished clips come back out of it. The steps never hand files to each other directly: they hand over a key.
Read the write-up →In front of the delivery layer
A bucket on its own is slow for a worldwide audience. Put a CDN in front of it and the same file is answered from near the viewer, with the bucket itself closed to the internet.
Content delivery and CDN →Managed and self-hosted together: hybrid
Because the S3 API is the same everywhere, this is the layer where hybrid is easiest to actually build. The working copy can sit on your own MinIO, the processed result in a cloud bucket, and the code that moves between them changes one endpoint. The decision stops being technical and becomes what it always was: which files may leave.
What would help you next?
Learn to decide
The course “From Managed to Hybrid” teaches where AI should run: what each layer costs, what you operate and where your data has to stay. It is in preparation, with a free live session first.
Linda MohamedAI and cloud consultant, AWS Community Hero, Vienna
Build it with me
Tell me what you want to build. In 30 minutes we check which of these ways fits your data, your rules and your budget, and whether AWS funding for a proof of concept can cover part of it. How AWS funding works →
You already know what you need
Book the workshop directly:
- AI Concept and Prototype workshops →
Architecture first, priced per service and tested on your data, then a prototype with a go or stop decision. - AI Discovery Workshop →
Sort your ideas into a scored shortlist, so you know which use case to build first.
Read more on ai-solutions.wiki
My open knowledge base, with longer comparisons, pricing notes and sources.
Sources
- Amazon S3 documentation. https://docs.aws.amazon.com/AmazonS3/latest/userguide/Welcome.html
- Azure Blob Storage documentation. https://learn.microsoft.com/en-us/azure/storage/blobs/storage-blobs-introduction
- Google Cloud Storage documentation. https://cloud.google.com/storage/docs/introduction
- MinIO documentation. https://min.io/docs/minio/linux/index.html
- MinIO, ai-solutions.wiki. https://ai-solutions.wiki/tools/minio/
- Azure Blob Storage, ai-solutions.wiki. https://ai-solutions.wiki/tools/azure-blob-storage/
Links checked October 2026. Product names and features change: check the linked pages before you decide.
