I have spent forty years establishing whether numbers are true.

Converting the code is the easy half.

Most organisations planning to retire a SAS estate cannot answer three questions with confidence: what is actually in there, what depends on it, and what it will genuinely cost to move.

The estate has usually grown over fifteen or twenty years. Documentation is partial or gone. The people who wrote it have left. Undocumented business rules have accumulated — someone patched a variable in 2011 to fix a reporting problem and never wrote it down. A hardcoded date sits in a macro nobody has opened since its author retired.

Conversion tooling will quote on line counts. Line counts do not tell you which lines feed a regulator-facing number, which variables have no equivalent in the target platform, which code is dormant and can simply be deleted, or where the reconciliation will fail six months in.

These programs rarely fail at translation. They fail at validation — at the moment an actuary, a risk committee or an external auditor asks how anyone knows the new numbers match the old ones. By then the budget is committed and the legacy platform has a decommissioning date.

Generative AI has sharpened this rather than solved it. Translation is now close to free, which means validation is a larger share of the remaining risk than it has ever been.

SAS Estate Discovery & Migration Readiness Assessment

Before you commit a budget, you get an accurate picture of what you are actually moving — and an honest estimate rather than an optimistic one.

It stands alone. Take the findings and run the migration internally, or put them to another provider, and the engagement has done its job.

1. Program and job inventory. Every program, macro, stored process and scheduled job, with size, last-modified date, execution frequency and probable owner — including the orphaned and dormant code, routinely a third of what is found.
2. Dependency and data lineage map. Source system through to final report. This is what tells you which parts can be retired immediately, which are load-bearing, and what breaks if moved in the wrong order.
3. Variable-level gap analysis. Which legacy fields have clean equivalents in the target, which need transformation, and which have none. Regulatory and financially significant gaps flagged individually.
4. Ranked risk register. Showstoppers named explicitly — configuration limits, undocumented business rules, licensing dependencies, single points of knowledge failure.
5. Costed migration options. Two or three viable routes with effort estimates, indicative cost ranges and timelines, including partial retirement where full migration is not justified.
6. Reconciliation and tolerance strategy. How equivalence gets proven, what tolerance is defensible, how many parallel-run cycles to budget, and who must sign off and on what evidence. This is the section that determines whether your program completes.
7. Executive summary and presentation. Two pages for the steering committee, plus a walkthrough delivered to your sponsor and stakeholders.

One accountable person, and the risk where it belongs.

Delivered by me personally. No subcontracting, no offshore component, no rotating team. Remotely from Sydney, in Australian time zones.

That last point is demonstrated rather than asserted: the most recent comparable program ran entirely remotely for a nationally distributed team including its directors, across seven months and fourteen parallel-run cycles, to external audit sign-off. Remote delivery keeps more of the budget on the work.

What you are buying is someone who can read the system nobody documented — and who knows which parts of a machine-generated conversion to distrust before an auditor finds them.

Engagements are structured so the risk sits where it belongs. Discovery is fixed price. Design and build run against fixed milestones, priced from the discovery findings. Parallel-run cycles are charged per cycle, because cycle count is driven by your reviewers and auditors — so you keep control of that spend, and I carry no incentive to extend it.

What buyers ask.

What does the assessment involve?

Four weeks producing a full inventory of every program, macro, stored process and scheduled job; a dependency and data lineage map; a variable-level gap analysis against your target platform; a ranked risk register; costed migration options; and a reconciliation strategy setting out how equivalence will be proven and who has to sign it off. It closes with a two-page executive summary and a walkthrough for your steering committee.

How much does it cost and what are the terms?

$48,000 plus GST, fixed. Four weeks from the date access is granted. Fifty per cent on commencement and fifty per cent on delivery, fourteen day terms. Not subject to variation unless scope is changed in writing.

What if our estate is much larger than typical?

The fixed price covers estates up to approximately 500 programs and macros, which covers most. Larger estates are quoted after a short no-cost sizing call — a 3,000-program estate is a materially different job, and pretending otherwise would mean either overcharging you or underdelivering.

Can NSW Government agencies engage Uncharted Worlds directly?

Yes. Uncharted Worlds Pty Ltd is an accredited supplier under the NSW Government ICT Services Scheme. For engagements within the relevant threshold, agencies may negotiate directly with an accredited supplier rather than going to market, and the assessment is priced with that in mind.

We already have a code conversion tool. What does this add?

Conversion tools translate code, and they do it well. They do not tell you which lines feed a regulator-facing number, which variables have no equivalent in the target platform, which code is dormant and can simply be deleted, or how you will prove to an auditor that the new outputs match the old ones. These programs rarely fail at translation. They fail at validation.

What access do you need?

Read access to the SAS estate — code, logs and metadata — and read access to the target platform schema where a target has been chosen. No write access is required or requested. Around six hours of subject matter expert time across the four weeks, in short sessions, and a named sponsor with authority to accept the findings.

Will you write any code during the assessment?

No production code, and no change to any production system. Analysis tooling is used to inventory and trace the estate, but the deliverable is the assessment itself. Its purpose is to let you commit to the next phase against an accurate estimate rather than an optimistic one.

What happens if we want to proceed to the build?

Delivery is priced after discovery, because until the estate is mapped nobody can quote it honestly. Design and build run against fixed milestones. Parallel-run reconciliation is charged per cycle. Implementation, post-go-live verification and handover are fixed.

Is the assessment fee credited against the build?

No, deliberately. If it were creditable you would reasonably wonder whether the recommendation had been shaped by the follow-on work. The assessment is designed to stand alone, so you finish it entirely free to take the findings to your own team or another provider. That independence is most of what you are paying for.

How many parallel-run cycles should we budget for?

More than most programs assume. A genuinely complex estate can run to double figures — the most recent comparable program took fourteen over seven months. Cycle count is driven by your reviewers and auditors rather than by development speed, which is why cycles are charged individually and you keep control of that spend.

What if we decide not to migrate at all?

That is a legitimate outcome and sometimes the right one. The options paper includes partial retirement, and where a full migration is not justified the assessment will say so. You will also have a documented estate, which is valuable in itself if the people who wrote it have left.

Do you work remotely?

Remote delivery from Sydney is the standard arrangement. The most recent comparable program ran entirely remotely across a nationally distributed team including its directors, for seven months, through fourteen parallel-run cycles to external audit sign-off. Work is performed in Australian time zones.

Who actually does the work?

Michael Charters, personally. No subcontracting, no offshore component and no rotating team. Forty years in enterprise and regulated data, SAS Certified, director of Uncharted Worlds Pty Ltd since 1995.

Which clients do you take on?

NSW Government agencies, where Uncharted Worlds is an accredited supplier under the ICT Services Scheme. Australian federal agencies, with prior delivery at the Department of Veterans' Affairs and the Department of Human Services. And regulated private-sector organisations — insurers, banks and superannuation funds — operating comparable SAS estates.

Are you insured?

Professional indemnity and public liability cover is in place, with certificates of currency provided on request. Uncharted Worlds Pty Ltd has been ASIC-registered and trading since 1995, ABN 39 071 410 707.