Assurance you can trace: every finding back to its evidence.
GA Applications designs compliance, quality and environment dashboards that organise your approved requirements, checks, findings, observations, corrective actions and review dates. Every measure shows its evidence, method and provenance, corrective actions have owners and due dates, and legal interpretation, certification and professional conclusions remain with qualified reviewers.
Assurance work fails quietly: a check that stopped happening, a finding nobody owned, evidence that cannot be found at audit time. This direction makes the chain from requirement to evidence visible — and makes gaps look like gaps.
Demonstration uses fictional requirements and checks. Real builds organise the requirements your organisation approves.
Assurance chain — requirement to evidence
Interface demonstration — fictional data-
REQ-114 · Workshop extraction maintenanceiRequirement v3owner: H&S coordinator · source: approved obligations register · review due 30 Sep
-
Checks this quarter: 13 scheduled · 12 done · 1 missed▲1 gapmethod: inspection form F-22 · evidence: 12 forms + 9 photos attached · missed check 14 Jul — no record, shown as gap
-
Finding FND-031 · filter overdue on unit 2■Openraised by: M. Henare, 18 Jul, form F-22 + photo · severity: agreed scale, level 2
-
Corrective action CA-118 · replace filter, verify airflow▲Due 25 Julowner: T. Kauri · closure requires: photo + qualified verification · history preserved after closure
-
Reviewer sign-off?Awaiting qualified reviewthe dashboard records evidence and state — the compliance conclusion belongs to the qualified reviewer
Built for people who answer auditors
This direction serves quality managers, environment and compliance reviewers, H&S coordinators and the executives accountable for assurance. Its decisions are traceability decisions: Which requirements have gaps in their check schedule? Which findings are ageing without an owner? Can we produce the evidence for that claim — who checked, when, by what method — in the next five minutes? The view organises what your approved obligations require and what your records show; whether that amounts to compliance is a conclusion for your qualified reviewers, and the design never pretends otherwise.
The target operating outcome: no silent gaps, no orphaned findings, and audit evidence retrieved rather than reconstructed — with the boundary between record-keeping and professional judgement drawn in the open.
The assurance chain, with provenance at every link
Requirement → check → finding → action → review
Every obligation enters as an approved requirement with an owner and a review date — not a copy of legislation, but the commitment your organisation has signed up to, versioned. Checks are scheduled against it, and the schedule itself is watched: a missed check is a visible gap, as the 14 July miss is in the demonstration, never quietly skipped. Findings raise corrective actions with owners, due dates and closure criteria. Closure requires recorded confirmation, and closing an action never erases the finding that caused it — the history is the assurance.
Provenance is the product
Each piece of evidence carries its origin: who performed the check, when, by which method, against which requirement version, and what form the evidence takes — a form submission, a photo, an imported result, a document. These are different weights of evidence and they are labelled differently, so a reviewer can weigh them instead of counting them. Where environmental sensors or monitoring equipment feed the view, readings carry source, calibration context and freshness like any other evidence; GA Applications designs the software layer, while equipment selection, calibration and statutory sign-off remain with qualified people.
Careful measures, not vanity metrics
Assurance numbers invite gaming, so the measures are chosen defensively: check completion against schedule, finding ageing by severity, action closure against due date, evidence retrievability. "Percentage compliant" is deliberately absent — a blended score would hide exactly the gaps this view exists to expose.
Review-date calendar — next 60 days
Interface demonstration — fictional data| Requirement | Review due | Reviewer | State |
|---|---|---|---|
| REQ-098 · Contractor induction | 31 Jul | H&S coordinator | ▲Evidence pack assembling |
| REQ-121 · Stormwater inspection | 15 Aug | Environment reviewer | ●On schedule |
| REQ-114 · Extraction maintenance | 30 Sep | H&S coordinator | ■Open finding blocks sign-off |
| REQ-130 · Ladder register | — | — | ◐Owner vacant — escalate |
A requirement with no current owner is an exception in itself — like REQ-130 — because unowned obligations are how schedules quietly die.
Definitions this view typically carries
Check completion vs schedule
Checks performed divided by checks scheduled per requirement per period. Missed checks remain visible as gaps — rescheduling does not erase the miss.
Finding age by severity
Days open per finding, grouped by your agreed severity scale. Severity definitions come from your framework; the view reports against them, it does not set them.
Corrective action closure
Actions closed with recorded confirmation divided by actions due in the period. Closure without the required evidence does not count as closed.
Evidence retrievability
Sampled requirements where full evidence chain (check, method, provenance, review) is retrievable within the agreed time. Measured honestly by sampling, not assumed.
Where evidence lives, and how it stays trustworthy
Sources
- Obligations register: approved requirements, versions
- Inspection forms & photos
- Imported results (lab, sensor) with context
- Findings & actions registers
Governed layer
- Requirement versioning & ownership
- Check scheduling & missed-check detection
- Evidence provenance: who, when, method, form
- Retention & integrity controls
Views
- Reviewer: requirement & evidence chains
- Coordinator: check schedule & gaps
- Action owners: own queue
- Executive: assurance summary (no blended score)
Evidence integrity is engineered, not assumed. Records are append-first: corrections are new entries referencing the old, with author and reason, so the trail survives scrutiny. Retention follows your obligations and is documented per evidence type. AI assistance can help summarise approved evidence for a reviewer, but it never generates findings, never alters records and never substitutes for the reviewer's conclusion — consequential determinations remain with the qualified humans accountable for them. If a source feed fails, affected chains show their evidence as incomplete with the gap named, because half a chain presented as whole is the worst outcome in assurance work.
What we need from you, and how delivery runs
Inputs
- Your approved obligations register — the requirements you track, with owners
- Your check methods, forms and where evidence currently lands
- Your severity scale, closure criteria and retention rules
- The reviewers who will use the chains, and the boundaries of their role
Delivery & acceptance
- Decision workshop: which requirements, which evidence, which reviewers
- Source profile: where evidence lives today and how provable it is
- Prototype one requirement end-to-end — schedule, check, finding, action, review
- Pilot across one review cycle with real reviewers
- Acceptance: missed checks visible as gaps; provenance complete on sampled chains; closure requires confirmation; corrections preserve history; no blended compliance score anywhere; table equivalents and keyboard access pass
Right fit
- You answer to audits, certification bodies or boards on quality/environment/compliance
- Evidence exists but is scattered across inboxes, drives and clipboards
- Findings and corrective actions lack owners and ageing visibility
Wrong fit
- You want software to certify compliance or replace qualified review
- You want health, clinical or legal advice generated by the system
- You want evidence history editable without a trail
Related directions and services
Frequently asked questions
Does the dashboard tell us whether we are compliant?
It shows the state of checks, findings and actions against the requirements you have approved. Whether that constitutes compliance is a conclusion for your qualified reviewers, auditors or legal advisers — the dashboard organises their inputs and evidence; it does not replace their judgement.
What is evidence provenance?
Each recorded check shows where it came from: who performed it, when, by what method, and against which requirement version. A photograph, a form submission and an imported lab result are different kinds of evidence and are labelled as such, so reviewers can weigh them appropriately.
How are corrective actions tracked?
Every finding can raise a corrective action with an owner, due date and closure state. Overdue actions are visible in the exception queue, closure requires a recorded confirmation, and the history is preserved — closing an action does not erase the finding that caused it.
Can environmental sensor data feed this view?
Yes, where monitoring equipment exists or is planned — GA Applications designs the software and integration layer, while equipment selection, installation, calibration and statutory approval remain with appropriately qualified people. Sensor readings show their source, calibration context and freshness like any other evidence.
How does this help at audit time?
The evidence is already organised by requirement, with provenance and review history, so producing it is retrieval rather than archaeology. Reviewers can drill from a requirement to its checks, findings and closures without someone assembling a folder the night before.
At your last audit, how long did it take to produce the evidence?
Tell us which obligations, standards or internal standards you track and where the evidence currently lives. We will scope a view that keeps requirements, checks, findings and corrective actions traceable — with the review boundary clearly drawn.