Why We Open-Sourced Shaide

When we started axem, we set out to solve a problem that most of the AI industry has quietly decided to ignore: Why should being regulated mean being left behind on AI? Defense organizations, aerospace primes, and regulated enterprises don't get to use the AI tools everyone else takes for granted. Their data can't leave the building. Their networks are air-gapped by law, not by preference. And yet they face the same pressure as everyone else to move faster, retain institutional knowledge, and put AI to work - just without the option of a SaaS login. We built shaide to close that gap. Today, we're taking the next step: we're open-sourcing its core under Apache 2.0.

Sovereignty isn't a feature - it's the whole point

"Sovereign AI" gets used as a buzzword a lot these days. For us, it has a very specific meaning: an organization should never have to trust a vendor, a foreign jurisdiction, or a cloud provider's uptime to keep its most sensitive knowledge under its own control. That's true whether you're a defense ministry, an aerospace manufacturer, or a hospital. But sovereignty has a second dimension that's easy to miss: you shouldn't have to trust a vendor's roadmap, either. Closed, proprietary infrastructure means that the moment a vendor changes pricing, gets acquired, or shuts down, the institution built on top of it inherits that risk. Open-sourcing shaide's core means the organizations who deploy it aren't just running it on their own infrastructure - they can read the code, audit it, fork it, and keep running it. That's what real sovereignty looks like, and we think it's a standard the whole industry should be held to, not just something we say in a pitch deck.

Giving back to the community that built this space

None of what we've built exists in a vacuum. shaide stands on the shoulders of years of open-source work in model serving, retrieval, and on-prem infrastructure - tools built by people who believed AI shouldn't require a hyperscaler contract to run. We've benefited enormously from that ecosystem, and it's time we contributed back to it directly instead of just building on top of it quietly. We also believe the developers building air-gapped and on-prem AI systems - often inside defense and aerospace organizations that are starting to build this capability in-house - deserve a serious, production-grade starting point instead of reinventing deployment, orchestration, and compliance tooling from scratch every time. That's exactly the gap shaide's open-source core is meant to fill.

What's in the open-source release

The Community edition includes the parts of shaide that we believe should never be locked behind a paywall:

• Self-hosted deployment on-premises or in your own private cloud
• The core orchestration and deployment engine
• The community feature set, free to run indefinitely

What stays closed - and why

We're not open-sourcing everything, and we want to be upfront about that. The enterprise features that require sustained engineering investment remain part of our commercial offering:

• The sensitive knowledge center with enterprise-grade user management
• Our chat agent
• Our regulated specific VS Code agent and CLI agent
• Dedicated support and custom feature development

That's how we keep building the open-source core in the first place. We think that's a fair trade: the foundation is free forever, and the parts that make Shaide enterprise-ready are what fund its continued development.

What this means if you're building in this space

If you're a developer or engineering team inside a defense, aerospace, or compliance-sensitive organization already building your own on-prem AI tooling, we'd rather you start from a codebase that's already solved the hard infrastructure problems than build from zero. Pull the repo, run it, break it, tell us what's missing. If it saves your team months of infrastructure work, it's done its job.

What's next

This is the beginning of a longer relationship with the open-source community, not a one-time announcement. We'll be maintaining the repo actively, and we're looking for feedback, contributions, and - frankly - pressure-testing from people operating in environments far more demanding than most software ever sees. You can find the repository on GitHub today: https://github.com/axem-solutions/shaide