ISH crown
ISH Commercial OSIsgcawu Somnotho Holdings (Pty) Ltd
CLIENT REVIEW COPY

REQUIREMENTS → EVIDENCE → DESIGN → VERIFICATION

An invoice workspace
you can inspect and trust.

The case for the improved ISH Commercial OS, with visible examples, source-backed pricing checks and clear acceptance boundaries.

Review edition: 9 September 2026 · Software candidate a9ed01e · Technology by Aleph Creative Hub

What this report establishes. The isolated upgrade has working interface and pricing-control improvements. Initial published-price component checks are complete. Full regional pricing accuracy, hosted course delivery and client acceptance remain open.

01 / THE CLIENT'S JOB

Make the next action obvious.

Create with confidence

Initial use was difficult. Creating an invoice needed a clear starting point and a reviewable path from the agreed quotation.

Control the whole invoice

The user needs to see the document, select its parts, adjust details and distinguish one invoice's selling price from future supplier costs.

Know where prices came from

Regional rates must match the product, unit, location and date. AI assistance must explain uncertainty and help with the active task.

These priorities come from the user's brief and the reviewed Ronald feedback. They are requirements, not a claim of completed client acceptance.

02 / THE RESEARCH

Local evidence first. Explicit comparisons where needed.

10Local material observations
8Published local delivery bands
1Matched Vryburg product
12/12Component checks passed

The data pack contains 17 observations: 10 local material prices, one comparison product, three civil-engineering wage floors and three broad building-cost references. All remain review data. The 10 local material observations come from one retailer; they are not ten independent supplier confirmations.

Observed local material prices, 9 September 2026. VAT included; delivery and handling separate. Leaflet observations expire 20 September 2026.
Product / specificationPrice unitPublished priceEvidence
AfriSam all purpose 42.5N, 50 kg50 kg bagR122.85Source
Sephaku 42.5N, 50 kg50 kg bagR118.45Source
Vera Grey 35 x 35 cm; box covers 2 m2m2R112.95Source p.1
Icon White porcelain 60 x 60 cm; box covers 1.44 m2m2R179.95Source p.1
Hercules Stone 42 x 42 cm; box covers 1.59 m2m2R122.95Source p.1
Square tube 50 x 50 x 1.6 mm, 6 m6 m lengthR357.95Source p.4
Square tube 75 x 75 x 1.6 mm, 6 m6 m lengthR534.95Source p.4
Lip channel 75 x 20 x 50 x 2 mm, 6 m6 m lengthR464.95Source p.4
Rectangular tube 76 x 38 x 1.6 mm, 6 m6 m lengthR399.95Source p.4
Dulux Rockgrip PVA white, 20 litres20 litre containerR629.95Source p.2

What the research changed

Choosing a comparison area

Vryburg is a candidate for the Kuruman-area material comparison: an inland store with the same retailer and exact product. Its semi-arid setting is documented in CSIR environmental assessment material. This is an analytical choice, not proof of equal construction costs. Compare supplier routes, freight distance, availability, labour access, project scale and specifications before using any substitute.

Do not use Vryburg as a universal substitute for Upington, Kimberley or the entire Northern Cape. Soil, water availability, access and mine-site requirements remain project-specific. AECOM provides broad cost context, with its stated Gauteng basis and exclusions retained.

Coverage still missing: independently corroborated bricks, bulk sand and stone, ready-mix, plant hire, installed trade rates and actual supplier-account terms. Sean Blake, Olivier and other local supplier quotations still need verified price evidence. No arbitrary regional uplift fills these gaps.

03 / SHOW THE CALCULATION

A real price, a clear quantity, a visible transport assumption.

Twenty AfriSam 42.5N bags are 1,000 kg. Using the observed Kuruman price:

20 × R122.85 material priceR2,457.00
20 × R1.00 displayed handling chargeR20.00
Illustrative delivery at 60 km from KurumanR1,725.00
Sum of displayed componentsR4,202.00

This is a benchmark illustration, not a supplier quotation or installed-concrete price. It assumes one delivery. Confirm handling-tax treatment, checkout terms, vehicle/load limits and road access. Published adjacent distance bands share endpoints; exact boundary distances require clarification.

The same AfriSam SKU in Vryburg was R113.65 per bag, compared with R122.85 in Kuruman: an observed difference of approximately 8.10% for this pair. It does not establish a category-wide or province-wide uplift.

Inspect the diagnostic findings

The existing fallback contains a 1.306 cement-category factor. A sensitivity probe using two repeated copies of the Vryburg ex-VAT value returns R129.07, compared with the observed Kuruman R106.83 ex VAT: 20.82% higher. Those copies are a synthetic probe, not two independent national sources. The result identifies a calibration risk, not measured end-to-end forecast error.

R122.85 ÷ 1.15, rounded to two decimals, becomes R106.83. Multiplying that rounded unit by 20 gives R2,136.60. Dividing the gross material line once gives R2,136.52: an eight-cent difference. The release needs an explicit precision policy; this diagnostic does not claim that policy has already been changed.

04 / THE INTERFACE DECISIONS

See the page. Make a change. Return with confidence.

A structured document canvas keeps invoice meaning intact while allowing section movement, zoom and direct selection. Contextual forms make precise adjustments accessible. The mobile revision gives Page, Edit and Layout separate views, larger controls and readable inputs.

Earlier mobile editor
Earlier mobile editor
Layout controls and small actions compete with the document. Actual 390 px capture.
Revised mobile editor
Revised mobile editor
Page, Edit and Layout separate viewing from adjustments. Same 390 px viewport and top-of-page position.
A deliberate price change
A deliberate price change
Scrolled detail view: unit, invoice-only scope, reason and source are grouped together.

Protect the user's work

Edits survive view changes. A selling-price change requires a reason. The saved state and preview remain visible. Undo supports recovery; issued-document snapshots preserve their recorded state.

Help in context

Lerato receives the active invoice context and can inspect it or propose supported changes. Read-only guidance is presented separately from actions that need review. Its collapsed mobile control leaves room for the task.

Short depth transitions support orientation. Reduced-motion styling is present. Responsive emulation is not proof of physical-device comfort or a complete accessibility audit.

05 / WHAT HAS BEEN CHECKED

Evidence with a defined scope.

AreaEvidenceLimit
Core upgrade427 backend tests at core candidate f35ba1e, including disposable Business records checks.Earlier core validation; not rerun for this presentation report.
Current mobile candidate178 tests across 48 frontend files and Next production build passed before commit a9ed01e.Local candidate. Production unchanged.
Mobile journey320, 360 and 390 px containment; 44 px header controls; draft retention; reason guard; invoice-only save and restore.Browser emulation. Physical phone and client acceptance pending.
Published-price components12/12 checks: ten VAT conversions, actual regional resolver SQL and missing-item behaviour.Captured persistence parameters and isolated isolated test records. Not live web extraction, assistant pricing or Business records acceptance.
AI tool behaviourEarlier real Lerato assistance responses for inspection, specified price correction and section movement passed server replay; mutations rolled back.Does not certify current price discovery or every unsupported request.
Quotation to invoiceFive synthetic source lines preserved; R11,500 + R1,725 VAT = R13,225.Demonstration data, not a payable customer invoice.
Teaching and brandingActual screenshots on each manual demo; two review lesson videos; full ISH name and small Aleph Creative Hub credit.Remaining lesson videos, human voice review and hosted delivery remain open.

A staging request to /api/contracts returned 404 during the mobile run. It is not silently counted as a clean whole-application smoke test. The tested invoice path and reported checks completed.

06 / THE CLIENT DEMONSTRATION

Show the outcome and the evidence beside it.

  1. Start from the agreed work. Select an accepted demonstration quotation, show the customer and total, and create one draft.
  2. Show control of the document. Select a section, edit a reference, move Payment notes and use Undo. Check that layout changes leave the total intact.
  3. Demonstrate a deliberate price change. Show the item unit and source. Change the demonstration selling price, supply a reason and save. Explain the separate future-cost choice.
  4. Open this benchmark example. Compare the exact product and store, show transport separately and explain why unverified items remain unverified.
  5. Ask for help. Request invoice guidance from Lerato. Show what it inspected and distinguish advice from a proposed action.
  6. Repeat on a phone-sized screen. Use Page, Edit and Layout, then show the saved preview.
  7. Open the learning and architecture materials. Demonstrate screenshots, practice tasks, review videos and architecture views. Demonstrate online feedback only after hosting is verified.
  8. Record acceptance. Capture Ronald's observations against the requirements. Automated checks support his decision; they do not replace it.

07 / RELEASE ACCEPTANCE

What remains before calling the work complete.

08 / INSPECT THE EVIDENCE

Sources, scope and decisions.

The accompanying data pack preserves URLs, observation dates, product units, VAT basis, store names, validity, source-page references and hashes of retrieved files. No benchmark row is approved for the company cost book.

Official company identity: Isgcawu Somnotho Holdings (Pty) Ltd. This client review uses synthetic interface records.