IronShard is S3-compatible object storage for AI agents, built on a decade of EU-funded storage research. Every file is encrypted and split into fragments across multiple storage providers.
We are a spin-off of Chocolate Cloud ApS, a Danish storage company. Our team's work there spans European research projects on encrypting data, erasure-coding it into fragments, spreading those fragments across cloud and edge locations, and deciding where data should live.
10 years
of storage research, since 2016
10
research projects
€2M+
in EU funding to the team
Seven EU projectsFrom 2016 to 2028
Horizon 2020 and Horizon Europe research, each with a public record in CORDIS, the European Commission's database of EU-funded research.
Secure big data processing in untrusted clouds, with TU Dresden and Imperial College London. The team ran its erasure coding inside Intel SGX enclaves, so fragments were created where the cloud provider could not see them.
A secure cloud continuum across edge, cloud, and HPC, with the University of Stuttgart and Mellanox Technologies. The team built an S3-compatible gateway that encrypts files, erasure-codes them into fragments, and spreads them across cloud and edge storage.
Machine learning for running systems across edge and cloud, with Inria, TU Delft, and NTT Data. The team built storage that uses AI to decide where data goes, and adapts as conditions change.
Automated security for software supply chains. The team brought its SkyFlok storage platform in as a use case, with vulnerability checks in the build pipeline and a public Trusted Bill of Materials.
AI-driven collaboration between IoT devices and edge resources, with NEC Laboratories Europe. The team builds one storage layer over resources owned by different organisations plus public cloud, through a gateway that keeps working offline.
A €28 million cloud-edge-IoT platform tested in eight real-world pilots, with 55 organisations including Intel, Schneider Electric, and Telefónica. The team supplies the S3-compatible object storage the pilots use.
A Marie Skłodowska-Curie training network on the Internet of Things, with Aalborg University, TU Eindhoven, and Telefónica. The team hosted two PhD researchers and developed compressed, network-coded IoT storage that can be queried in place.
Components and subsystems for faster, lower-energy communications infrastructure for IoT, 5G, and Industry 4.0, with DTU, Aarhus University, and twelve Danish companies. The team led a task.
A pilot with Partisia and Aarhus University that combined the team's network-coded storage with multi-party computation, keeping IoT data protected in storage, in transit, and in use.
Energy-efficient, secure communication infrastructure, from optical fibre and satellite links to 5G and IoT security, with DTU and 18 Danish companies, including KMD. The team is a project partner.
SecureCloud (grant agreement No 690111) and SERRANO (No 101017168) received funding from the European Union's Horizon 2020 research and innovation programme. MLSysOps (No 101092912), ReSCALE (No 101120962), EMPYREAN (No 101136024), and O-CEI (No 101189589) are funded by the European Union under Horizon Europe. Chocolate Cloud ApS is the beneficiary in each of these six projects. IoTalentum (No 953442) received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie actions. Chocolate Cloud ApS took part as a partner organisation.