Amazon Web Services (AWS) said on April 8 it has launched S3 Files, combining file system access with its S3 object storage.

Previously, file-based tools could not directly access S3 data, requiring data to be copied to a separate file system or complex synchronisation pipelines. S3 Files addresses the issue.

S3 Files is built on Amazon EFS and automatically translates file system operations into S3 requests. Applications can access S3 data without code changes, and the data does not leave S3. Thousands of computing resources, including instances, containers and functions, can connect simultaneously to the same S3 system, while access via the existing S3 API is maintained. It uses existing buckets as they are, so data migration is not required.

AI agents can maintain and share state across pipelines, and machine learning teams can run data preprocessing without preparing or copying files separately. Analytics teams can directly access data lakes with file system tools. S3 Files caches frequently used data to provide low latency and supports read throughput of several terabytes per second.

S3 Files is available immediately in 34 AWS regions, including Frankfurt, Zurich, Stockholm, Milan, Spain, Ireland, London and Paris.