One framework. Every stage of a governed agent.
Agentic AI that isn't governed is a liability. K9-AIF makes governance part of the architecture, and the K9X products cover every stage around it: design, build, verify, run, decide, and stay current.
The framework, and the ecosystem that keeps it current
Layers that know only the one below. Governance, Zero Trust and people across every layer. Every box opens what it names.
GuardianGovernanceevery agent
in and out signed
identityZero Trustverify pause,
ask,
resumeHuman in the loopa person
decides trace
eventsAudit and traceevery
decision PLATFORM SERVICES · THROUGH FACTORIES, SWAPPABLE Kafka event busPersistence and routing stateObject storage (S3 / MinIO)MCP toolsIdentity: OIDC (Keycloak, Entra ID, Okta)Secrets
The lifecycle
Every stage has a tool. Every tool is built on the framework it serves.
What each one does
Every product is built on K9-AIF itself.
Design a solution visually and generate a framework-compliant scaffold to build on.
Open → K9X ContinuumCatalog of building blocks by TOGAF level, with review and promotion of reusable solutions.
Open → K9X Enterprise RepositoryArchitectures, standards and decisions in one place.
Open →Checks solutions on every push: contracts, layering, governance, secrets and framework version, with a remediation guide.
Open → K9X SatanFires real attacks at a governed pipeline to show what the framework's controls contain and where gaps remain.
Open → K9X SentinelReads new AI-security research daily, judges each threat against what the framework covers, and raises gaps for review.
Open → K9X NewswatchWhat shipped in your AI stack, and what to check: vendor releases and docs, verified against the source, scored on one governance rubric. By email if you like.
Open → K9X ArenaModels compete on real tasks; graded results become evidence for the framework's model routing.
Open →Human-in-the-loop review: agents raise decisions, people make them, and the outcome flows back to the running solution.
Open → K9X Streams & k9x OriginLive, governed data for agents: one call returns what is true right now, where it came from, and any reason not to trust it.
Open → K9X DashboardOne view of the ecosystem: what is running, what is waiting for a person, and where things stand.
Open →YouTube
| Video | Length | Link |
|---|---|---|
| K9-AIF DAS in depth: one requirement, end to end | 6:43 | Watch ▶ |
| Learn agentic AI development: free course and certification (K9X Developer Foundation) | 2:46 | Watch ▶ |
| Enterprise Repository & Continuum: know what you have, reuse what works with K9X | 2:56 | Watch ▶ |
| Secure AI Lifecycle: security from inception to production with K9X | 2:59 | Watch ▶ |
| K9-AIF Patterns: the Event Router | 1:52 | Watch ▶ |
| K9-AIF Patterns: one path to any model | 2:26 | Watch ▶ |
| K9-AIF Patterns: how components are built and wired | 1:46 | Watch ▶ |
| K9-AIF Patterns: integration | 2:04 | Watch ▶ |
| K9-AIF Patterns: how the work gets done | 1:50 | Watch ▶ |
| K9-AIF: Zero Trust explained | 1:56 | Watch ▶ |
| K9X Sentinel and Satan: how K9-AIF keeps up with AI-driven attacks | 2:45 | Watch ▶ |
| k9x Origin: AI agents don't run on yesterday's data | 2:01 | Watch ▶ |
| K9-AIF: the framework in 3 minutes | 2:51 | Watch ▶ |
| DAS: agents prepare, humans decide | 4:26 | Watch ▶ |