Twenty-three gaps and ten decisions. Take a position on each; the notes matter more than the ticks.
A single figure with the arithmetic attached — what was measured, where, by which instrument, with what uncertainty, and which companies had reported by the time it was struck.
A register that holds every claim line against the approved work programme and the contract terms, and puts a human auditor in front of the exceptions instead of the pile.
An asset inventory covering ministry-operated facilities: what equipment, which vendor, which firmware, in support or out of it, and the record of remote access to it.
An independent Iraqi measurement of Iraqi production, produced before the secondary-source panel publishes, with the divergence from that panel attributed field by field.
A baseline: every custody and allocation point, its class, its uncertainty, its last proving date and its next one — on screen, including the gaps.
A standing record of what was drilled, built or flared within a defined corridor of a shared structure, and when — held by Iraq, timestamped, and independent of anyone's report.
A per-contract record assembled as the contract runs — approvals, variations, claims, acceptances, payments — rather than reconstructed from files when it is asked for.
A mapping layer that takes each company's chart of accounts as it stands and produces one consolidated view — without asking any company to change its system.
A single identifier issued at clearance and carried on the bill of lading, the certificate of origin and the manifest, so a cargo is the same object at every step.
A repository with a published submission specification, a confidentiality window, and a three-way split between who owns the data, who administers the repository and who operates it.
A monthly, per-stack record placing observation beside declared utilisation with the tolerance of each, so a divergence becomes a case with a date on it rather than an argument.
A structured register of contract terms — baselines, fees, R-factors, performance factors, cost-recovery ceilings — so a quarterly settlement can be checked against the instrument rather than against a recollection of it.
Reconciliation of depot loading against delivery receipt, with the tank level watched in between — the same pattern as hydrocarbon accounting, one link further down.
A zone-and-conduit model per facility with the target security level set against the level actually achieved. The gap between them is the finding.
A register that shows every item awaiting signature, which authority band holds it, and how long it has been there — so that a delay is a fact on a screen rather than a rumour.
Arabic document intelligence over the correspondence and contract archive — extraction, classification and retrieval with a citation back to the page.
A defined process: notification received, classified, operator notified, response recorded, outcome verified, result published — each step with a window and an owner.
One asset register on a published data model, with linear referencing so an inspection finding, a repair and a risk score all land at the same point on the same centreline.
The reporting path itself: a case record that classifies an incident against the national scoring standard and produces the report inside the window that standard sets.
The reconciliation that transparency reporting requires, produced as a by-product of the system rather than as an annual manual exercise.
A single view of major projects against the capacity, throughput or export route each one unlocks — so a slipped commissioning date is visible in the production plan it affects.
A gateway per body — signed, authenticated, logged — so a query is a recorded transaction rather than a letter, and nobody surrenders a database to anyone.
A portfolio holding every deployed model with its owner, its measured precision, the cases it touched, and a retirement record for the ones that did not work.
These are the ministry's to make. They are here because the answer changes what we build, and because it is better to ask now than to discover later.
The UK regulator publishes named-operator compliance under a standing policy. Whether Iraq does the same, and when, is a ministerial decision. The instrument prepares the card either way.
A scorecard that excludes the state companies is not a scorecard. This is the single item most likely to be quietly removed after approval, so decide it now, in the open.
Part of this answer is not software. Nigeria audited 187 flow stations before building anything. Without it, the screens display precise wrong numbers.
The code is mandatory in its own words but was never gazetted. Law 84 of 1985 already empowers the Minister to specify what is reported and how often. That is an instruction, not new legislation.
We do not ask any company to change its system. We ask that what it already produces arrives in an agreed shape at an agreed frequency. That is an internal decision before it is a technical one.
Notices already arrive. Where no focal point is designated they go to a diplomatic inbox. This one costs a letter.
Reconciliation proves itself on a few hundred trips before it expands. Route optimisation does not work in year one and we do not propose it.
Year one covers ministry-operated sites only. Anything inside a contractor's facility is a contract conversation, not a technical one.
Iraq has no data protection statute and no supervisory authority. Handling commitments therefore have to be contractual. Pretending otherwise would be a compliance claim with nothing behind it.
No institution absorbs ten new systems in a year. We would rather deliver three that work than promise eleven that do not.