Government — Poolside
Poolside is a U.S.-based AI company delivering frontier open-weight models purpose-built for on-premises and air-gapped deployment—giving government agencies full sovereignty over their AI infrastructure, data, and costs.
Unlike cloud-based AI services, there are no per-token fees: agencies run unlimited inference on their own hardware, making Poolside a predictable, scalable investment.
- Air-Gapped and Disconnected
- STIG-Hardened OS
- ATO Achieved
- Cleared Resources
- IL5 Deployable
- On-Prem/GovCloud/Edge
Who we work with.
"Partnering with Poolside allows Vibrint to deliver cutting-edge AI capabilities into the most sensitive government environments without compromise.
Their technology is purpose-built for the security and performance requirements of federal missions, and our joint work demonstrates how innovation and operational security can advance together.
This partnership is enabling solutions that directly support warfighter readiness and national security mission success."
Vibrint
Christopher T. Oliver, General Manager—Products
"Security isn't negotiable for our customers in the public sector. Poolside being airgapped by default is what makes the partnership work—our customers get best-in-class AI-assisted development without compromising on the requirements that matter most to them.
That's why it's a core part of our full-stack offer."
Sterling Computers
Andrew Peppler, AI Strategic
"Poolside has fundamentally changed how software is written for secure compute environments. Poolside models have not only changed how we build secure software internally for our customers at Hunted, but being a Poolside partner enables us to provide the American warfighter the best models for software development that are used to secure software in support of critical national security missions."
Hunted Labs
Hayden Smith, CTO & Co-Founder
Most AI companies build for the cloud and add security later. We built Poolside security first.
From day one, our stack was designed for environments where cloud-only platforms cannot go—because we believe the most consequential work in the world happens with an organization’s critical data, within their security boundary.
Secure, by design
Air-gapped and on-premises deployment are the foundation for all of our architectural decisions, not capabilities retrofitted later. Every query, every agentic workflow, every response runs entirely within your infrastructure. No external calls. No data egress. No shared systems. Your most sensitive data and missions stay exactly where they belong—inside your boundary.
Own your intelligence, permanently
When you deploy Poolside, you receive full model weights—not just API access, not a rate-limited cloud endpoint, not a vendor-controlled subscription. You control the weights, how it runs, and how it evolves. No dependency on Poolside's infrastructure. No risk of a vendor pulling access. Your AI capability becomes a permanent, sovereign asset.
Successful outcomes
We don't hand off models and walk away. Our experts—cleared to work in government environments—ensure our models and inference engine are deployed the right way and ready to scale.
- Reduced delivery timelines
- Improved mission system reliability
- Increased productivity in secure environments
- Agentic pipelines capability with full audit trails
The AI foundation deployed entirely inside your boundary.
Every capability, every model, every agentic workflow runs within your infrastructure from day one.
Foundation Models
Purpose-built for software engineering. Poolside models deliver multi-step reasoning, autonomous task execution, and complex software engineering workflows optimized for on-premises inference across all supported hardware configurations. You receive full model weights. No cloud dependency, ever.
One foundation
We deliver the intelligence foundation: models, inference, and agents. That foundation is platform-agnostic by design, so it powers your people wherever they work, whether that's an engineer in an IDE, an analyst in a chat window, or a custom application calling the API. You bring the surface. We bring the intelligence underneath it.
Powered by Atlas
Atlas is Poolside's inference engine, and it improves model performance by up to 30%. That means running our models in your own environment delivers more value on the same compute, so you get higher output from existing infrastructure while keeping full control of the models you run.
Choose the deployment model that fits your environment.
FULL STACK
The complete Poolside stack inside your boundary. Foundation models, inference, infrastructure, and security—delivered as a single integrated system.
MODULAR
BRING YOUR OWN HARDWARE
AVAILABLE ON AWS
Poolside doesn't just deliver a model. We show up.
Our team includes cleared personnel and solutions architects ready to operate inside the environments our customers work in every day.
Solutions Architects
Poolside SAs work with your teams from scoping through deployment, ensuring the models and inference are configured to your requirements.
With deep experience in hardened STIG environments, our SAs navigate security, compliance, and operational constraints; connecting AI with your data sources.
Cleared Personnel
Poolside maintains cleared personnel capable of supporting classified programs and operating within secure facilities.
Bring AI capabilities into the environments where your most critical mission work happens, without compromising security posture or access controls.
In the open
We're committed to releasing our models as open weights, because governments need a credible open frontier lab built on shared values.
You run our models in your own environment, adapt them to your mission, and stay operational regardless of what happens to any single vendor.
Bring AI inside your mission today. Talk to our team about deploying Poolside.
First name
Last name
Company
Your company email
How can we help?
I agree to the Privacy Policy