IA_AUT_ECO is a French Deep Tech initiative built on 45 years of technical expertise and 20+ years of Unix/Linux systems engineering. We provide a sovereign, full-stack infrastructure—from the hardened operating system to federated analytics—designed to empower people with disabilities through high-security robotics and ethical AI.
Winterhold is the secure foundation of our ecosystem. As a specialized fork of FreeBSD, it provides the kernel-level isolation and deterministic execution required for life-safety autonomous agents.
- Security: Native integration of Capsicum and Jails for agent sandboxing.
- Resilience: Real-time priority management via the Hearthfire kernel module.
- Integrity: Zero-Trust architecture enforced at the driver level.
IA_AUT_ECO is committed to the highest standards of open-source supply chain management. We align our operations with the OpenChain Specification (ISO/IEC 5230:2020) to guarantee legal transparency and security.
- Infrastructure (Ivarsread): Licensed under Apache License 2.0 for institutional interoperability.
- OS Core (Winterhold): Hardened FreeBSD fork following the original permissive BSD license structure with proprietary security enhancements for the commercial layer.
- AI Assets: All deep learning models follow an OpenChain-compatible traceability matrix (Licenses: MIT, Apache 2.0).
- SBOM (Software Bill of Materials): We maintain a comprehensive inventory of all third-party components and neural weight origins.
- DCO (Developer Certificate of Origin): All contributions must be signed-off (
git commit -s) to certify code provenance. - Vulnerability Management: Strict adherence to our
SECURITY.mdprotocol for coordinated disclosure.
Candidate for LF AI & Data (Linux Foundation), Ivarsread is a federated analytics server providing institutional-scale insights while maintaining 100% data residency on-premise.
The proprietary user interface layer, providing a secure and accessible bridge between the end-user and Winterhold-powered agents.
- DAWNSTAR: Autonomous mobility (OSM-native).
- ESBERN: High-precision robotic manipulation.
- VESTA: Behavioral anomaly detection for patient safety.
- Edge-Only AI: No sensitive data ever leaves the local Winterhold environment.
- Federated Learning: Model improvements are aggregated anonymously via Ivarsread.
- AI Act Readiness: The Ysgrammor protocol ensures full causal traceability for every autonomous decision.
IA_AUT_ECO operates under a technical meritocracy. We invite collaboration from experts in:
- Hardened Unix kernels (FreeBSD/HardenedBSD).
- Federated Learning and Edge AI optimization.
- ISO/IEC 5230 compliance and Open Source legal engineering.
Founder: Polyglot Programmer (C/C++, Python, SQL) | Unix Expert since 1977. Location: Meaux, France. Contact: contact@iaauteco.ovh
