A practical guide to the public beta tools, what advisory and unverified results mean, why a registered surveyor must sign off, and how to try everything with synthetic sample files — no real client data needed.
Five steps to try the public beta safely. Every result stays advisory and unverified.
Download a fictional .rpt / .csv / evidence file — no real client data needed.
Findings are labelled ADVISORY / NEEDS REVIEW / UNVERIFIED — never final.
Step through OK / WARN / FAIL / REF / CHECK findings with the evidence beside each one.
Export an advisory review summary — a working file, not a survey deliverable.
When you are ready, request a pilot for your firm's workflow.
Pick the tool that matches what you have. Each runs on synthetic or cleared test files during public beta and produces advisory, unverified output only.
Not sure? Start with the documentation or the supported files register. Every tool stays advisory and unverified — not for survey deliverables; registered surveyor confirmation required.
SurveyMarkup is in public beta. Everything it produces is advisory and UNVERIFIED — it is a checking aid, not a survey deliverable and not a legal or cadastral certification.
Use the tools to speed up review, but a registered Australian surveyor must confirm any result before it is relied on.
The QA demo runs SurveyMarkup's deterministic checks on a fixed synthetic survey sample — no upload and no account needed.
Edit the project/survey metadata, run the checks, and read the advisory findings. AI explanations are generated from the engine output shown unchanged.
Map Linework runs advisory data-quality checks on coded geometry — MGA zone validity, CRS/datum, missing codes, open polygons, duplicate/zero-length geometry, closure, RL jumps, pole height, prism constant, GNSS quality and resection geometry.
It produces an export-readiness checklist. "Blocks export" means a data-integrity problem, not a legal judgement; exported CSV/GeoJSON must still be reviewed.
Report Reader reads plain-text .rpt and .txt survey reports, detects the encoding, and extracts labelled metadata, warning/error/misclosure/residual lines and a parse-confidence score.
Load the built-in synthetic sample or drag a file in. Low parse confidence is flagged as a partial parse; review the original report carefully.
PDF Markup links evidence/markup items to QA findings and shows an evidence summary: linked, unlinked, needs-review and missing items, plus measurement review and reviewer notes.
Real PDF upload is in pilot; the public demo runs on a synthetic evidence set. Exports are advisory only and are explicitly not a compliance certificate.
The supported-files register lists what SurveyMarkup can import, review and export today, what is in pilot, and what is not supported (for example, OCR for scanned PDFs and any legal/cadastral certification).
Use it as the authoritative file reference, not as a promise of downstream compatibility.
Every result is labelled. ADVISORY means an informational hint. NEEDS REVIEW means a reviewer should look more closely. UNVERIFIED means a tolerance/measurement result has not been confirmed against the applicable standard.
Nothing in SurveyMarkup is "verified" on its own. Results become trustworthy for deliverables only after a registered surveyor confirms them.
SurveyMarkup does not determine whether a survey satisfies regulatory or cadastral requirements, and it never sets a result as surveyor-confirmed on its own.
A registered/licensed surveyor must review and sign off before any deliverable, lodgement, or legal/cadastral decision. This boundary is intentional and is not bypassed in the beta.
AI explanations are clearly labelled and are generated from the engine's own output, shown unchanged. The AI does not invent numbers or change tolerance results — numeric and compliance guards stay on.
AI usage is rate-limited and capped. Treat AI text as an explanation aid, never as a verification.
Please do not upload confidential or real client data during the public beta. Use the synthetic sample files below to try every tool safely.
Where parsing happens in your browser (Report Reader), your file is read locally; in all cases your original file is preserved and never modified.
We are designing a workflow to track a registered surveyor's review — status, reviewer details, the rule source/version/effective date, unresolved items, evidence and comments. Why sign-off is required: SurveyMarkup is a checking aid and cannot make the professional call.
What the current beta does: it shows advisory, unverified findings only. What the future workflow may support: structured review tracking and an audit trail. What stays manual: a registered surveyor's professional judgement and the sign-off itself.
It is a design preview only — it tracks workflow status, never confirms a rule, and never sets surveyor_confirmed=true. SurveyMarkup never self-certifies.
SurveyMarkup previews Australian coordinate context — the GDA2020 and GDA94 datums, MGA zones 49–56, the AUSGeoid2020 and AUSGeoid09 geoid models, and AHD/RL heights — and flags pairings that look inconsistent (for example a datum and geoid model that are not conventionally used together).
These are advisory metadata-plausibility checks, not coordinate authority. SurveyMarkup does not transform, recompute or verify coordinates, datums, zones or heights. A flagged mismatch is a prompt to look more closely, never a determination that something is right or wrong.
Because the checks read declared metadata, a value can be present and self-consistent while still being the wrong choice for the actual data. Every result stays UNVERIFIED, and a registered surveyor must confirm the correct datum, zone and geoid model against the project brief and applicable standards before reliance.
The /tools section hosts eight free advisory calculators and validators — COGO (bearing, distance and area), grid ↔ ground (combined scale factor), traverse misclose, level loop closure, a point file validator, a coordinate and datum converter, a deliverable checklist generator and a datum/metadata quick-check. Every tool shows the full working for each calculation so each step can be re-checked.
Where a published guideline applies, results only ever read “within guideline” or “exceeds guideline — review”, and guideline values are read from the published advisory rule packs. Every result is advisory and unverified — not a survey deliverable — and a registered surveyor must confirm it independently before reliance.
Every major check explains what it inspects, the formula or heuristic it uses, the default threshold and where that threshold comes from, and its false-positive / false-negative limits. Thresholds are triage defaults that need registered-surveyor review — nothing here is a pass/fail conclusion, and every result stays advisory and unverified.
These explainers describe data-quality triage only. They do not certify a survey, do not confirm a coordinate result, and do not replace registered-surveyor judgement. A registered surveyor must validate every threshold and confirm every finding before reliance.
Try every tool without uploading real client data. Each file is fictional, carries the PUBLIC BETA — ADVISORY ONLY note, and produces advisory, unverified output only.
A preview of the future sign-off workflow. It tracks workflow status only — it never confirms a rule, never sets surveyor_confirmed=true, and never creates a survey deliverable.
Design preview only — this workflow does not confirm any survey rule, does not set surveyor_confirmed=true, and does not create a survey deliverable.
Final compliance assessment requires a registered Australian surveyor. SurveyMarkup does not determine whether a survey is legally compliant or non-compliant. Engine results are checking aids only; a registered/licensed surveyor must review and sign off before any deliverable, lodgement, or legal/cadastral decision.
Placeholder fields — no surveyor is assigned in this preview.
A future registered-surveyor confirmation step would live here. It is intentionally inert: there is no control that sets surveyor_confirmed=true, and the audit note records that no confirmation was performed — “No confirmation performed in this design preview; surveyor_confirmed stays false.”.
Wording you will see across SurveyMarkup: advisory, needs review, unverified, registered surveyor confirmation required, and not for survey deliverables. SurveyMarkup never claims a survey is certified or compliant.
Advisory, unverified walkthroughs for common Australian survey review workflows. Every output needs registered surveyor confirmation and is not for survey deliverables.
Plain-language definitions of the Australian survey terms SurveyMarkup works with. SurveyMarkup previews and surfaces these signals for review — it does not transform, verify or certify coordinates. A registered surveyor confirms anything before it is relied on.
SurveyMarkup does not certify survey accuracy, cadastral compliance, legal boundaries, or coordinate correctness. It helps prepare and review working information, but original data and final outputs must be checked by a qualified user before use.
Assistance only - professional review remains required.