Meta-orchestrator, not a replacement.
Works with your existing Slurm, Kubernetes, and Flux stack. No migration, no retraining.
The operator already has a scheduler, a storage fabric, and a team that knows how to use them. Amalgamy does not replace any of that. It sits above the existing stack as a meta-orchestrator, adding intent translation, facility-aware scheduling, and sovereign compliance enforcement. The operator's tools keep working. The researchers keep their notebooks. The infrastructure gets smarter.
What it integrates with
Three layers, zero replacements.
Amalgamy is a meta-layer that sits alongside existing investments. Every integration is additive. Nothing is removed.
Schedulers and orchestrators
Amalgamy sits above Slurm, Kubernetes, LLNL Flux, and Mirantis k0rdent. It does not replace them. It adds intent translation, facility awareness, and compliance enforcement on top of the operator's existing scheduler.
Silicon and hardware
Unified, hardware-agnostic support for NVIDIA, AMD (ROCm), Intel, and emerging accelerators. The operator runs mixed-silicon environments without per-vendor scheduling scripts.
Storage and data fabrics
Native awareness of high-performance storage solutions: Weka, DDN, VAST, and Lustre. Amalgamy calculates the network cost of data movement and routes accordingly.
What the operator gets
Keep the stack, gain the platform.
No rip-and-replace.
The operator's current schedulers, tools, and workflows keep working. Amalgamy makes them smarter, not redundant.
No retraining.
Researchers work in Python, Jupyter, Bash, Airflow, MLFlow, Nextflow, and standard IDEs. No new syntax, no new tools.
No vendor lock-in.
NVIDIA, AMD, and Intel silicon in the same fleet. The scheduler handles the translation. The operator picks hardware by workload, not by scheduler compatibility.
Laptop to cluster in one step.
A model running locally transitions to a partner cluster by changing the data input, not by rewriting the script.
Storage-aware scheduling.
Amalgamy knows where the data lives and calculates whether to move the data or move the task. Eliminates the I/O bottleneck that makes allocated GPUs idle.
Future-proof portability.
New schedulers, new silicon, new storage fabrics. Amalgamy's meta-layer means the operator adopts new infrastructure without rewriting their orchestration.
Deploy over existing infrastructure, not instead of it.
Technical docs at amalgamy.ai. Engagement and capacity planning at ThisWayGlobal.