Today we’re announcing the Poolside Platform: a production-grade system for running AI agents inside your boundary, anywhere your business needs it to run. The Poolside Platform is built to help enterprises adopt AI while not compromising on security or governance.
One platform to manage AI with the most complex requirements. You choose the model and the Platform meets you where that choice lands.
Most AI systems today are something you rent with token fees you can't control, based on third-party API access. A Poolside deployment provides you with complete control of the model weights and a predictable subscription plan. With full visibility, auditability and telemetry. An enterprise grade built for those that run the world’s most important business.
Who we built this for
We built the Poolside Platform for a specific kind of team: the kind whose approved base images are a matter of policy; whose network architectures make sure no single node has both GPU access and outbound connectivity; whose CISO has been forced, over the past two years, to relax controls for an AI vendor because there wasn't a real alternative. Now there’s Poolside.
Inside the Platform, agents run where developers already work—VS Code, Visual Studio, Zed, IntelliJ, or the terminal—and connect to the tools teams actually use, like Slack, Jira, GitHub, Workday, and Salesforce through centrally managed MCP servers. Every session runs in a containerized environment with built-in secret management and network policy control. Every agent action, including tool calls, file edits, reasoning steps, and decisions, is recorded as a searchable trajectory. Credentials are encrypted at rest, injected at runtime, and automatically redacted from outputs. Administrators define the boundaries centrally; the Platform enforces them.
Four ways to deploy: on your own bare metal, air-gapped on turnkey hardware through partners like Dell; inside your existing VPC on AWS, Azure, or Google Cloud.
And Forward Deployed Research Engineers who actually ship it with you. Our FDREs sit with your team, work out where agents create the most value in your specific workflows, build on the Platform alongside you, ship the first automated workflow in weeks, then harden and measure.
There's more than we can cover here. Learn more on our enterprise page, which goes deeper on the Console, sandboxed execution, audit retention, and the deployment modalities.
Why now
The fast-moving AI coding market has progressed primarily through cloud APIs requiring enterprises to relax enterprise safeguards and controls, or allow dangerous modes of operating to avoid falling behind. Businesses deserve better. AI vendors should not require enterprises to relax well established best practices, and Poolside does not. Enterprises expect AI auditability and governance to become mainstream priorities in 2026 [], pushed there not by regulators but by the people responsible for running AI in production.
What this enables
For any team whose data is too sensitive, too regulated, or too strategic to send outside of their security boundary, the Poolside Platform puts AI within reach.
For a bank that's been quietly running an AI coding tool under a three-month security exception, with Poolside, you no longer need to compromise security for your team's efficiency. For a defense contractor whose AI usage policies have been running ahead of the tooling available to enforce them, it means your controls can finally match your requirements.
We're excited about what customers have already built with the Platform. Hunted Labs, who helps you protect the security of your software supply chain, has made Poolside part of its full-stack offer:
Poolside has fundamentally changed how software is written for secure compute environments. Poolside models have… changed how we build secure software internally for our customers at Hunted”
That's the story we want to tell more of. Not because the demos will be flashy, though some of them will be. Because the workflows that run in these environments are, quite literally, the ones that matter most.