Four phases. Four pillars.
Three principles.
The operating system behind every Rubix engagement. Phases run sequentially. Pillars run in parallel through every phase. Three principles govern the work. A library of frameworks does the lifting.
The phases sequence the work. Each phase has a deliverable. Each deliverable funds the next.
Sequenced. Each phase has a deliverable.
Engagements start where the client is. Clients with a strong internal diagnostic skip Frame and enter at Strategize. Clients ready to ship enter at Build. The phases sequence the work, not tax it.
00
Frame. Diagnose. Align. Decide.
We diagnose what the client actually has, align senior stakeholders on the question worth answering, and produce an honest call on whether the work is worth doing. If it isn't, we say so. That is part of the value.
Duration
2–4 weeks
Our approach
01
AI Maturity diagnosis. Five-level scoring on data, governance, talent, technology, and process. We see where the client really is, not where they hope to be.
02
Senior stakeholder alignment. Working sessions with the executive sponsor, function heads, and IT/data leadership. Surface the real question, not the framed one.
03
Value-chain quick-scan. First-pass map of where AI is plausible and where it is not. We exit Frame with a candidate list, not a fantasy.
04
Engage / pause / decline call. An honest recommendation: ready for Phase 01, not yet, or never. Most consultancies cannot say the third option. We can.
What you receive
Frame Document: the diagnostic on a page
AI Maturity baseline: the 5-level scorecard
Engage / pause / decline recommendation
Optional: scoping for Phase 01
What to expect
2–4 weeks, mostly on your premises. We need access to people, not data. The data work begins in Phase 01.
Complimentary. The Frame deliverable is yours regardless of what happens next.
Phase 00 is how we earn the right to Phase 01. If our diagnosis isn't compelling, you should not engage us further.
Decision gate at the end. Proceed to Strategize? The only commitment we ask for at this point.
Frameworks applied in this phase
Phases run in sequence. Pillars run in parallel, through every phase, without exception.
Parallel. Every phase stresses all four.
The pillars are how we know the engagement is sound at any phase. A Build that ignores Governance & Risk produces software that cannot go live. A Strategy that ignores Change & Adoption produces a roadmap that cannot ship. Each pillar is the audit dimension of the work.
Pillar 01
Governance & Risk
Risk and governance stitched in from Phase 00, not appended at Phase 03. International standards because Saudi enterprises are increasingly audited against them.
NIST AI RMF mapping per use case
ISO/IEC 42001 governance baseline
Eval-driven release discipline
Bias and fairness as release blocker
Audit trail with provenance
Pillar 02
Data & Knowledge
AI is only as good as the corpus it stands on. Data work is architecture, not preparation; institutional memory is a first-class asset.
Data architecture diagnosis
Retrieval-layer design
Institutional memory layer
Lineage, provenance, access control
Curated bilingual AR/EN corpora
Pillar 03
Technology & Platform
Platforms designed to survive a model swap, a vendor change, a regulatory shift. Sovereignty is a default architecture decision.
Multi-model orchestration
Sovereign / KSA-resident inference
Existing-system integration
Edge / cloud / hybrid deployment
LLMOps lifecycle discipline
Pillar 04
Change & Adoption
The hardest part of an AI engagement is not the model. It is the operating model. Built for client capability and ownership from day one.
Hub-and-spoke CoE design
70-20-10 capability building
Operating-model before software
Adoption metrics tracked
Hand-the-keys exit posture
Phases describe how we work. Pillars describe what we audit. Principles describe what we will not do.
Three principles. Non-negotiable.
They hold whether we are running a 2-hour workshop or a multi-year transformation. They are the discipline that keeps every Rubix engagement honest. If we cannot apply them, we decline the work.
Principle I
I.
Process before platform.
No tool conversation until the underlying process is mapped. Tools are chosen by the process they serve, never the reverse. This protects clients from buying technology that solves the wrong problem.
What this rules out
Vendor demos before diagnosis. Tool selection before value chain mapping. "AI-first" thinking.
Principle II
II.
Evidence before enthusiasm.
Every use case must be tied to a measurable business KPI: time, cost, error rate, compliance, leakage, revenue. If it cannot be measured, it does not get built. Conviction is not a substitute for a business case.
What this rules out
Pilots without baselines. ROI claims without numbers. "Innovation theatre."
Principle III
III.
Readiness before roadmap.
Data, governance, talent, and change capacity decide what is deployable inside twelve months. We say no to what will fail, so the portfolio delivers what will win. Ambition is disciplined by readiness.
What this rules out
Roadmaps the organization cannot absorb. Bold plans on broken foundations. Strategy as wishful thinking.
In practice
When these three principles meet a client engagement, some ideas survive intact, others are reshaped, and a few are set aside for later. That filtering is the value we add, before any AI is deployed.
The frameworks below are not decorative. Each one gets named where it is applied.
The consulting library behind the work.
Every Rubix engagement draws from a defined set of frameworks. Some are international standards we apply. Others are Rubix-developed and refined across engagements. They are not decorative. Each one gets named where it is applied.
Every engagement walks the same path.
Eight ventures in our portfolio walk this path. Every engagement in our archive walks this path. The case is the receipt.