You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
To develop and document a dual-node 'Shield and Vault' hardware architecture that allows for high-utility local AI inference without cloud-based data exfiltration.
1. Successful deployment of an air-locked RTX 5090 inference node.
2. Publication of an open-source 'Sovereignty Manual' for independent hardware assembly.
3. Zero-telemetry verification of the internal network.
* Hardware/Inference Node: $10,000 (RTX 5090 Tier).
* Life Stabilization & Environment: $7,500 (Covers transitional "life debt" and a neurodivergent-conducive workspace to ensure 100% focus).
* Living Stipend: $33,000 (6 months of full-time research buy-out).
* Tax Reserve (Est. 35%): $18,000.
The $10,000 minimum represents the core hardware procurement cost (Inference Node). Reaching this threshold allows for the technical validation of the 'Shield and Vault' architecture, even if the living stipend and stabilization goals are deferred.
Project Aegis is currently led by Damocles, an independent systems architect and technical generalist. I bring a decade of multidisciplinary experience in high-stakes environments—from agricultural equipment repair and industrial maintenance to culinary and custodial management. This 'real-world' background ensures that Project Aegis is built with a focus on mechanical reliability, structural integrity, and practical usability rather than just theoretical code.
While Project Aegis is my first formal entry into the public open-source AI safety space, my track record is defined by a career of system diagnostics and mechanical recovery.
Infrastructure Maintenance: 2+ years of experience repairing complex agricultural feed and manure handling systems, where I developed a 'Zero-Failure' mindset toward hardware.
Technical Foundation: Completed the core Computer Science curriculum in a focused academic setting before transitioning to field-based engineering, giving me a unique 'First-Principles' approach to logic.
Project Management: Successfully managed logistics for high-output environments (culinary and custodial), specializing in process efficiency and contamination prevention—principles I am now applying to digital data sovereignty.
1. Hardware Procurement & Volatility "The most likely cause of a project delay is the volatility of the high-end hardware market. Specifically, the procurement of the RTX 5090 inference node and the custom server architecture.
Outcome: If specific hardware becomes unavailable or exceeds the budget due to market shifts, the project will pivot to the next most efficient 'Open-Sovereignty' hardware stack (e.g., dual 4090s or optimized Mac Studio clusters). The research goal of creating an air-locked 'Shield and Vault' remains the same even if the specific engine changes."
2. Technical Integration Complexity "Creating a truly zero-telemetry, air-locked bridge between the Shield and the Vault is a high-difficulty networking task. There is a risk that the initial hardware configuration creates too much latency for a smooth user experience.
Outcome: If the 'perfect' air-lock proves too slow for real-time companionship, the project will document these friction points as 'Lessons in Latency.' We will then release a 'Hybrid-Sovereignty' protocol as an interim step, allowing the community to choose between maximum speed and maximum isolation."
3. If the Project "Fails" Completely "If the specific software/hardware bridge I am building proves impossible for a single architect to maintain at scale:
Outcome: The project is not a loss. All funding spent on hardware remains an asset of the research, and all documentation—including the 'Sovereignty Manual' and my 'First-Principles' audit of local LLMs—will be published to GitHub. The failure of the product becomes a contribution to the body of knowledge in AI safety, showing other independent builders exactly where the 'walls' are.
Amount Raised: $0
Sources: N/A
Additional Context: "Project Aegis has been entirely self-funded to date, including the initial research, architectural design, and the establishment of the open-source repository. This is the first formal request for external funding to transition the project from the conceptual design phase to hardware procurement and full-scale development."