Identity
Access and role control stay inside the appliance
Identity and access review are evaluated with the same system that runs AI.
GOVERNANCE · DEVRUN · EVIDENCE
Operate AI Execution and Project Control Together
Inside the IRON MIND appliance, local AI execution, DevRun approval, stage gates, audit evidence, delivery packages, collaboration messaging, and SIEM incident detection run in one operating flow.
NIS 12 Controls≈94 / 100IDE 3Alert Rules 30
Control Layer
The solution server handles project control inside the appliance
The solution layer is not a separate SaaS tool. It is the control brain that sits inside the appliance.
Identity and access
Role separation, SSO, and project access stay aligned with the control model.
Evidence capture
AI runs, artifacts, outputs, and operator context are stored as evidence.
Control points are placed inside the operating surface
Run AI locally
AI use starts inside the appliance instead of an external API path.
Capture context
Each AI action is linked to artifacts, operators, and project history.
Review before release
Governance and gate checks happen before submission.
Policy Tier
AI usage intensity should be separated by project conditions
Open internal use
Lower-risk work can move with lighter review while still staying inside the appliance boundary.
Controlled delivery work
Project-critical work keeps evidence and approval checkpoints visible.
Restricted work
High-risk work stays under stronger review conditions and tighter submission control.
Review what people actually decide in the operating surface
Issue management
Control decisions become meaningful when issue handling stays connected to project execution.
- Priority and owner
- Workflow linkage
- Delivery impact
AI DevRun
The control layer records AI use as project evidence rather than a disconnected experiment log.
- Prompt/output history
- Changed artifacts
- Review state
Governance board
Governance review is visible inside the operating surface instead of a separate spreadsheet.
- Exception queue
- SLA alerts
- Policy posture
Artifact management
Submission control only works when versions and review state stay visible in the same system.
- Artifact inventory
- Version history
- Review state
Evidence Fields
Evidence fields stay close to project work
- Prompt and output history
- Changed artifacts and versions
- Role and project context
- Review and approval history
AI runs board
Shows what was asked, generated, and changed in the project.
Governance board
Shows which exceptions and policy gaps still block the next step.
Gate board
Shows which milestone still lacks approval and why.
Artifact and report view
Shows what is ready for review, submission, or audit.
Next Step
Operate AI Execution and Project Control Together
IronMind v1.2.2. Product implementation completeness about 94 / 100, external certifications/audits passed: 0, NIS 12 controls plus mapping/delegation, SIEM, ARIA-256-GCM, mTLS, and 32-container air-gap installation verification.