GLOSSARY · AI STRATOPS
What is AI StratOps?
AI StratOps is the discipline that connects an organisation's AI ambition to its execution: which AI initiatives to commit to, in what order, with which partners and under what governance — decided at the level where the decision actually sits, then run to a steady operating cadence. It answers the questions an executive committee and a board ask about AI — not the questions an engineering backlog raises.
AI StratOps is not MLOps or AIOps.
MLOps industrialises machine-learning models: deployment, monitoring, retraining. AIOps applies AI to IT operations: incident detection, capacity, automation. Both are engineering disciplines, and both answer to the technology function. AI StratOps works above them. It does not ask how to run a model; it asks which AI initiatives deserve the company's capital and attention, in what sequence, built by whom, and governed by which decision rights.
·
MLOPS · AIOPS
AI STRATOPS
QUESTION
How do we deploy, run and monitor models and systems?
What do we commit to, in what order, with whom, and who decides?
OWNER
CTO, CIO, engineering and data teams
CEO, executive committee, board
UNIT OF ANALYSIS
Models, pipelines, infrastructure
Portfolio of initiatives, sourcing choices, decision rights
OUTPUT
Reliable systems in production
An arbitrated portfolio, a sequenced roadmap, a governance the board can follow
The four decisions AI StratOps settles.
01 — Where does value sit? The economics of the industry, and where AI shifts them — before any tool is chosen.
02 — What do we commit to, and what do we stop? The portfolio arbitrated on impact, feasibility and risk: commit, scale, pivot, wait or stop.
03 — Do we make, buy or partner? Each capability sourced on explicit criteria and five-year total cost, independently of any vendor.
04 — Who decides, and on what evidence? Decision rights, steering bodies, usage policy and the narrative the board follows.
From decision to operating cadence.
A strategy that is decided once and never revisited becomes a document. AI StratOps installs the rhythm that keeps it alive: quarterly portfolio reviews, stage-gate criteria, stop thresholds, and execution partners directed on the company's behalf. The method follows four movements — Frame, Decide, Create, Measure — and hands measurement over to AI FinOps, so that every commitment is later audited against objectives set before it began.
Who needs AI StratOps, and when.
Chief executives and executive committees who own the AI agenda. Boards asking who carries AI risk and what the programme is worth. Operating partners standardising AI across a portfolio of companies. Chief data and AI officers who need a mandate the business will follow.
It is required when the AI portfolio has grown faster than anyone governed it, before a budget cycle allocates next year's resources, when a structuring vendor or build decision looms, and when the board or the EU AI Act calls for clear decision rights.
Frequently asked questions.
How does AI StratOps relate to AI FinOps?
They are the two halves of one discipline. AI StratOps decides what to commit to and how it is governed; AI FinOps establishes what it is worth — the business case, the price that captures the value, and the return actually realised.
Who should own AI StratOps in a company?
The chief executive, with the executive committee. The technology function executes and advises; the decisions about value, sequence and risk belong to the business and to the board.
Is AI StratOps a one-off strategy exercise?
No. It starts with a set of decisions, then installs the cadence that keeps them current: quarterly reviews, stage gates and stop thresholds. AI moves too fast for a strategy that is written once.
Does Odyssey implement what it recommends?
No. Odyssey sells no development, integration or resourcing. When execution requires it, Odyssey directs the partners on the client's behalf rather than becoming one of them — which is what keeps a make, buy or partner decision independent.
Where does an AI StratOps engagement usually start?
Most start with AI Strategy Readiness, a scored diagnostic delivered in three business days, which establishes whether a portfolio arbitration, a sourcing decision or a governance blueprint should follow.