Sovereign AI and interactive HPC: unifying training inference and exploration in one workflow

Most AI teams run training on one platform and inference on another. The handoff between them happens outside any workflow definition—it's manual, fragile, and repeats with every model iteration. For organizations that can't send proprietary data to external AI services, this fragmentation isn't just inconvenient. It's a blocker.

In this webinar, we demonstrate Fuzzball service endpoints—a capability that unifies AI training and inference in single, portable workflows. This is sovereign AI without the complexity of managing separate platforms.

• Jonathon Anderson, Principal HPC Product Engineer, CIQ
• David Godlove, Product Engineering
• Eric Hendricks, Technical Product Marketing Manager

RESOURCES:

Fuzzball product page: https://ciq.com/products/fuzzball/ Fuzzball Documentation: https://ui.stable.fuzzball.ciq.dev/docs/ Request a demo: https://ciq.com/company/contact-us/

Press Release

CIQ delivers turnkey sovereign AI with service endpoints capability, transforming Fuzzball into complete training and inference platform

Blogs

• AI workflow orchestration: why separate platforms fail
• How to run interactive HPC workloads alongside batch jobs in a single workflow
• Running Jupyter notebooks on HPC clusters without SSH tunneling