Supreme Model T‑X
Sovereign AI Platform Foundation for UK Deployment
OrkinosAI Labs is building a UK-sovereign AI control plane — operator-controlled data, UK-controlled model weights, governed API access — designed for regulated and public-sector deployment. This page is the complete funding-readiness briefing for assessors and investment partners.
The UK's sovereign AI gap
Frontier AI development is concentrated in a small number of US and non-UK jurisdictions. UK public-sector and regulated organisations face data residency, audit, and accountability constraints that offshore-hosted models cannot reliably satisfy.
Model weights, training data, and inference infrastructure exist outside UK sovereign control, creating systemic risk for critical applications in government, defence, health, and financial services.
The strategic gap
- No domestically-developed UK AI platform offers a credible sovereign alternative with transparent governance controls.
- Procurement pathways for regulated AI require auditability, accountability, and demonstrable data-boundary controls — absent from mainstream offerings.
- UK organisations need UK-controlled weights, reproducible training, governed API access, and auditable deployment lifecycle.
What is Supreme Model T‑X?
A two-layer sovereign AI platform foundation: model development under UK control, governed API integration for regulated deployment.
- PyTorch training stack
- Dataset handling & provenance
- Evaluation pipelines
- Checkpoint lifecycle control
- Reproducible run configs
- OpenAI-compatible API surface
- Policy gating layer
- Operator-issued access keys
- Audit signal capture
- Rate and quota control
Data Sovereignty
Operator-managed training datasets with explicit provenance intent. No mandatory external data dependency.
Model Sovereignty
UK-controlled checkpoints. No mandatory hosted model provider dependency.
Infrastructure Sovereignty
Model core designed for hardware and cloud independence from inception.
Access Sovereignty
Operator-issued API keys and project-level access control. No shared credential risk.
Audit Sovereignty
Auditable usage records and deployment lifecycle signals for governance evidence.
Current maturity
Engineering progress evidenced across baseline runs, benchmark pipelines, CI operations, and governance documentation. All evidence is maintained in a reviewer-accessible state.
Baseline runs recorded
Initial model and evaluation runs completed with reproducible command and artifact references.
Benchmark pipeline
Configured and operational with maintained repeatability documentation for comparative testing.
CI/CD pipeline
Automated build, test, and deployment workflow operational and actively maintained.
Governance documentation
Architecture, evaluation, delivery, and risk documentation maintained in reviewer-ready state.
API governance layer
Policy gating intent, access control design, and audit signal structure defined and in development.
GPU compute
Provider options evaluated; procurement pathway documented. Activation pending funding confirmation.
Production hardening
Scoped to Phase 3 (Days 61–90). Dependent on Phase 2 GPU benchmark completion.
Pilot deployment package
UK-hosted infrastructure target selection in progress. Scoped to funded delivery phase.
Technology Readiness Level
Supreme Model T‑X is assessed at TRL 3–4: analytical and experimental proof-of-concept validated in a controlled environment with structured evaluation, reproducible run evidence, and operational CI/CD infrastructure.
- TRL 3 (achieved): Model training and evaluation workflows validated with baseline run records and documented artifact lineage.
- TRL 4 (in progress): CI/CD pipeline operational; governance layer and API access controls under active engineering development.
- TRL 5 (funded gate): Relevant environment validation — GPU benchmark cycle execution with comparative reporting. Contingent on funding confirmation.
- TRL 6 (funded gate): System prototype demonstration — UK-hosted pilot configuration, production hardening, and stakeholder demonstrations.
Honest posture: All claims are based on documented, evidenced work. Where deliverables are in-progress or funding-contingent, this is declared explicitly.
Delivery execution plan
Stabilise and prepare
- CI stabilisation and baseline consolidation
- GPU readiness planning and runbook updates
- Architecture documentation completion
- Governance evidence packaging for submission
- Risk register review and dependency tracking
Baseline evidence pack published; GPU provisioning pathway confirmed
GPU benchmark cycle
- Funded GPU provisioning activation
- Benchmark cycle execution against defined baseline
- Model checkpoint publication with provenance records
- Evaluation pipeline hardening and repeatability verification
- Pilot candidate packaging initiated
Benchmark report published; comparative baseline documented; pilot pack in draft
Harden and demonstrate
- Production hardening and reliability baseline
- Pilot-readiness packaging complete
- Stakeholder demonstrations with governance evidence
- Deployment pathway documentation for UK-hosted contexts
- Model card and audit evidence publication
Pilot-ready deployment package; governance evidence pack; stakeholder demonstration completed
Funding dependency: Phase 2 (GPU benchmark cycle) and Phase 3 (production hardening) are contingent on confirmed funding. Phase 1 deliverables proceed under current baseline without GPU dependency.
Governance, safety and auditability
Built-in accountability designed for regulated contexts from inception — not retrofitted compliance.
Public sector and regulated deployment
Central Government
Policy analysis, document processing, knowledge retrieval
Data residency, audit, classification boundary control
Defence and Security
Controlled AI workloads with restricted data
UK-controlled weights, no external model dependency
NHS and Health
Clinical decision support, administrative automation
Patient data boundaries, GDPR, clinical governance
Financial Services
Regulated analytical and compliance workflows
FCA-aligned audit, explainability, version control
Legal and Professional
Document review, research augmentation
Privilege and confidentiality controls, provenance
Risks and mitigations
Mitigation: Phase 1 designed to proceed without GPU. Alternative provider options evaluated; procurement pathway documented and ready to activate on funding confirmation.
Mitigation: Phase 1 deliverables proceed on current baseline. Phase 2 and 3 gate criteria explicitly scoped to funded activation.
Mitigation: Scope positioned as sovereign-first, not frontier-replacement. Use cases selected for domain fit. Transparency over overclaiming.
Mitigation: Multiple UK data-centre and private cloud options evaluated. Selection deferred to funded phase to avoid premature commitment.
Mitigation: Governance architecture designed for auditability. Aligned with current DSIT and UK AI Safety Institute guidance.
Mitigation: Configuration-driven run management and structured artifact storage mitigate reproducibility risk.
Risk register is maintained and reviewed at each delivery phase gate. Escalation pathways are defined for critical blockers.
Measurable outcomes at 90 days
Reproducible baseline
All baseline runs reproducible from published configuration files with documented artifact lineage.
CI pass rate
100% CI pipeline pass rate maintained as a delivery gate criterion across all active branches.
Benchmark report
Comparative benchmark report published within 60 days of GPU provisioning confirmation.
Model checkpoint
UK-controlled model checkpoint published with provenance and audit record.
Governance controls
Policy gating, access boundaries, and audit signals specified and reviewed before any external deployment.
Pilot configuration
UK-hostable pilot deployment configuration packaged and reviewed within 90 days of funding confirmation.
Technical evidence
All primary evidence is maintained in a reviewer-accessible state on this site. A structured technical briefing is available on request to walk through all materials in detail.
POC status · TRL assessment · Gap analysis · Funding readiness plan
Current proof-of-concept status: what is complete, what is in-progress, and what is funding-contingent.
Technology readiness level assessment against standard criteria with evidence references.
Current capability gaps, dependencies, and planned resolution pathway.
Investment case, funding use by phase, expected outcomes, and strategic return.
The complete evidence bundle is available to assessors and investment partners on request. Contact the team and we will send the current package immediately.
Request Evidence BundleReady to invest in UK sovereign AI?
We are actively seeking funding partners, public-sector collaborators, and sovereign AI programme leads. This submission is complete and available for structured technical review.