Meet PaletteAI

The software platform for deploying modern enterprise AI at scale.

See it in action

Escape DIY, without the lock-in

It’s a tale as old as time: do you choose the chaos and toil of building your own infrastructure, DIY… or opt for a turnkey platform that gives you simplicity, but limits your choice?

With PaletteAI, we give you the best of both worlds, and none of the tradeoffs. You get an enterprise-grade experience, with all the ‘batteries included’ comforts you expect: from a polished user interface, to prebuilt templates, to 24x7 support.

But you also get the freedom to build your AI factories the way you want them: on your choice of hardware, with your choice of ecosystem software.

AI factories ecosystem

A day in the life

PaletteAI is a software platform that serves both platform teams and AI/data scientists. 

Platform teams

Platform teams use PaletteAI Studio to assemble infrastructure and AI app stacks that meet policies and governance standards, saving them in a reusable repository for AI teams to access. Of course, we provide a host of profiles in our catalog for you to access out of the box.

Data scientists

Data scientists enter the platform, selecting and customizing the profiles that best fits their workload requirements. PaletteAI takes care of scheduling the workloads to the cluster of GPUs and DPUs provisioned to meet their needs.

The platform team keeps full control of observability, cost, security, scaling and utilization and can keep the infrastructure patched and healthy, so workloads run uninterrupted.

AI architecture diagram

A true enterprise platform

PaletteAI has been engineered to serve the most demanding enterprises and public sector organizations, featuring all the governance controls and toil automation you need to thrive at scale.

Simplified AI platforming

PaletteAI Studio
Design, observe and manage workload templates

Declarative AI stacks
Build full deployment models from infra to apps

AI framework support
RunAI, ClearML, NeMo, more

Version & lifecycle
Track changes, updates, rollbacks

Governance and security

PaletteAI Secure
FIPS Compliant, multi-tenant platform

Policy guardrails
Quotas, limits, cost insights

Zero trust
Identity and access enforced

Regulated compliance
For PubSec and regulated industry support

Separation of roles

Platform team controls
Create reusable stack templates for AI teams

Deployment guardrails
Workload federation, controls for tenancy

AI team self-service
Approved template deployment no IT toil

Role-based ops
Granular team permissions

Cost and resource efficiency

Automated provisioning
Dynamic allocation of AI hardware

Staff augmentation
Repeatable stacks reduce workload

Integrated observability tools
Built-in monitoring across clusters and workloads

Resource optimization policies
Quota, autoscaling, and right-sizing

This means we help you coordinate shared and dedicated resources, isolate teams and workloads, and manage access through granular RBAC.

Ready for your mission

PaletteAI is available in two editions: PaletteAI, for typical enterprise workloads, and PaletteAI Secure, for regulated industries and government use cases.

PaletteAI Secure builds on our heritage at Spectro Cloud serving high security industries, including federal government and defense.

It features FIPS 140-3 validated cryptography, and is backed by 24x7 support and SLAs.

Palette AI secure logo

Deployed where you need it

We know that every organization is different — so PaletteAI adapts to your requirements.

For the fastest time to value, we offer a shared, multi-tenant SaaS by default, giving you access to a management plane that we run and maintain for you.

If you need stronger isolation, we offer a dedicated SaaS instance for an additional charge.

We also make it easy to self-host your own PaletteAI instance in your own cloud or data center.

And if your mission demands it, you can deploy PaletteAI directly in an isolated ‘air-gapped’ site, where it can deploy AI workloads without the need to access the outside world.

SaaS
Multi-tenant & dedicated
Self hosted
Air-gapped
Price for value graph

Priced for value

Your AI initiatives change all the time: new team, new applications, new projects. So we’ve kept it simple: PaletteAI is priced based on a flat fee per GPU managed, including technical support.

So when you need to onboard a new team of data scientists, or reallocate resources to a major new training project, you don’t need to worry about the impact on your budgets.

Backed by the best

When you choose PaletteAI, you get access to the AI and infrastructure experts from Spectro Cloud, an award-winning provider trusted by organizations from the US Army to GE HealthCare.

But that’s not all. You also get the backing of a whole ecosystem of partners, including leading system integrators, managed service providers, and hardware vendors who can deliver a complete, validated solution with all the services you need.

See PaletteAI in action

The best way to understand the value that PaletteAI offers is to get a demo with one of our experts. Get in touch to start the ball rolling.

Meet us

Frequently asked questions

Does NVIDIA endorse PaletteAI?

Sorem ipsum dolor sit amet, consectetur adipiscing elit. Nunc vulputate libero et velit interdum, ac aliquet odio mattis. Class aptent taciti sociosqu ad litora torquent per conubia nostra, per inceptos himenaeos. Curabitur tempus urna at turpis condimentum lobortis. Ut commodo efficitur neque.

Is Spectro Cloud new to AI?

Sorem ipsum dolor sit amet, consectetur adipiscing elit. Nunc vulputate libero et velit interdum, ac aliquet odio mattis. Class aptent taciti sociosqu ad litora torquent per conubia nostra, per inceptos himenaeos. Curabitur tempus urna at turpis condimentum lobortis. Ut commodo efficitur neque.

Does NVIDIA endorse PIs PaletteAI different from Palette?aletteAI?

Sorem ipsum dolor sit amet, consectetur adipiscing elit. Nunc vulputate libero et velit interdum, ac aliquet odio mattis. Class aptent taciti sociosqu ad litora torquent per conubia nostra, per inceptos himenaeos. Curabitur tempus urna at turpis condimentum lobortis. Ut commodo efficitur neque.