Starfish Storage

Starfish Storage, founded in 2014, builds a metadata-driven platform for managing unstructured data at scale. Its software is aimed at file-based workloads in high performance computing (HPC) and AI/ML, where large volumes of files and objects must be classified, catalogued and moved across diverse storage systems. The company operates globally and reports managing well over an exabyte of capacity for its clients.

The platform rests on two main components. A metadata-rich catalog indexes unstructured files and objects, enabling automated data classification and the enforcement of metadata-driven storage policies across heterogeneous environments. Alongside it, a scale-out data mover and batch processor supports automated data movement and file-processing workflows, with user self-service and policy-driven management spanning both departmental file shares and the largest supercomputing file systems.

The approach is deliberately vendor-agnostic: the software works with hardware, software and cloud platforms including VAST Data, Weka, DDN, Hammerspace, Dell EMC PowerScale, IBM Spectrum Scale, Lustre, and S3-style object stores from AWS, Azure and Google Cloud. Reported use cases include archiving, data protection, migrations, cloud bursting, cost accounting, data disposition, ROT cleanup, FAIR data management and AI/ML workflows.

Starfish Storage's clients include 8 of the world's top 10 pharmaceutical firms, 7 of the 8 Ivy League universities and Department of Energy supercomputing sites. The company also serves organisations in semiconductors, oil and gas, fintech, automotive, healthcare, consumer products, and media and entertainment.