Built, shipped, and demonstrable.

Valtrin is young, and this page carries no borrowed logos and no rounded-up numbers. What it shows is built, in use, and demonstrable in a conversation — the standard every future entry here will meet.

01 · Estimating engineDemonstrable
02 · Operations platformIn production

Case 01 · Purpose-built software

The estimating engine: generate the quotation instead of rebuilding it.

The context. A fabrication quotation is an engineering document. Behind every price sits a weight calculation, a material build-up, labour hours, surface treatment, NDT and margin — assembled by a qualified engineer, usually in a spreadsheet, usually under deadline.

The problem. Each quote is rebuilt from the last one, edited in place, priced partly from memory. A revision request restarts the work. The pricing method lives in one file and one head — and when REV C differs from REV B, nobody can say exactly why.

The work itself is not difficult. It is repetitive, precise, and unforgiving of a mistyped decimal — exactly the shape of work a purpose-built tool should carry.

The spreadsheet way
Open the enquiry PDF
Retype dimensions into Excel
Look up weights by hand
Assemble costs line by line
Format in Word, export, email

5 manual stages · again at every revision

With the engine
Enter the scope as line items
Weight, material, labour computed
Quotation & BOM generated
Revision stored with signed deltas

1 entry task · the arithmetic is the system's job

What Valtrin built

A purpose-built estimating application. Quotation, weight, BOM and labour build-up are computed from published engineering data — ASME pipe schedules, WN flange masses, structural section tables. Pipework is priced by diameter-inch, plate by weld-metre banded on thickness, with grade multipliers for stainless and duplex. Every rate is editable, because in a real build the rates are yours. Nothing is a lookup of a finished price — the cost builds up the way a shop builds it.

There is no AI anywhere in the arithmetic. Every number comes from deterministic code: same input, same answer, every time — which is what makes the output auditable, and what makes the validation below possible.

Key functions

Quotation generation

Enter the scope once; the quotation assembles itself in the shop's own format, ready to issue.

Σ

Weight & material cost

Computed from published data — pipe schedules, flange masses, section tables — as each line item is typed.

BOM generation

The bill of materials falls out of the same line items. There is no second document to keep in step.

Labour build-up

Fabrication hours assembled the way a shop actually costs a job — not a single blended rate hiding the detail.

Revision comparison

REV B to REV C with a signed delta on every affected row. The spreadsheet can't tell you why. This can.

Editable rates

Material, labour, treatment, overhead, margin — every rate is yours to change, and the whole estimate follows.

Validation

The engine is validated against known results: feed it jobs whose outcomes are already trusted, and it must reproduce them — same inputs in, known totals out. The demonstration moment is a revision: switch a manifold from SS316 to duplex and lengthen it, and the weight falls while the price rises, with a signed delta on every affected row explaining exactly why.

100%deterministic — same input, same answer
REV B→Csigned deltas on every change
0invented numbers — validated on known results

Outcome

Quotation becomes an entry task instead of a rebuild. A revision becomes one change instead of an afternoon. The pricing method survives outside one person's spreadsheet — written down, in code, with a history that explains itself. The engine is real and demonstrable: ask for a walkthrough in the first conversation and watch it price a job live.

The interface

The Valtrin estimating engine — click the numbered points to explore
The job, identified

Number, client, enquiry date — the register starts here, not in a filename.

Scope as line items

Plate, pipe, sections, flanges — weights and material cost computed from published data as you type.

The cost builds itself

Material, labour, treatment, NDT, overhead, margin — assembled the way a shop actually costs a job.

Revisions with receipts

REV B to REV C shows every delta, signed. The spreadsheet can't tell you why — this can.

Click the numbered points

A real Valtrin-built tool — not a mockup.

Case 02 · Systems & portals · In production

A four-role operations platform, live in production.

The context. A scheduling-heavy education operation: sessions, rooms, teachers, students, parents — all moving, all needing a different window onto the same facts. The client's business is not named here; the system is demonstrated live in conversation instead.

The problem. This class of operation usually runs on a spreadsheet and a messaging app. The schedule has no way to refuse a clash, so clashes surface on the day. Attendance is reconstructed after the fact. And four kinds of people — owner, teacher, parent, student — all ask their questions of the same overloaded humans.

Without a system
Schedule kept in a spreadsheet
Clash discovered on the day
Attendance marked on paper
Parents ask for updates one by one
Records reconciled at month end

Every answer passes through a person

With the platform
Session entered once, checked against every commitment
Each role sees its own live view
Attendance on clock-in
Records stay current by themselves

One dataset · four windows onto it

What Valtrin built

A complete operations platform, built end to end: database schema, row-level security, authentication, role permissions, scheduling engine, deployment, domain and search indexing — handed over live. Four role-separated portals — owner, teacher, parent, student — sit over one dataset, roughly ninety pages in all. Each role sees exactly the view its work needs, and nothing it shouldn't.

Key functions

Conflict-prevention scheduling

Rooms, teachers and students are checked against every existing commitment at the moment of entry — the clash is refused before it exists.

Four role-separated portals

Owner, teacher, parent and student views over the same live data, separated by the permission model — not by copies.

Attendance on clock-in

Recorded at the moment it happens, on Malaysian time — not reconstructed at month end.

Documents & announcements

Flow through the same permission model as everything else, so the right people see the right notice.

Enrolment & administration

Students, classes and staff are entered once and appear everywhere they belong.

Permissions throughout

Row-level security under every query — the separation between roles is enforced by the database, not by good intentions.

Validation

Fifty-seven automated tests run against the platform and pass in production. They exist for the same reason the system does: rules like conflict prevention have to be right every single time, so they live in tested code rather than in anyone's carefulness.

4separated roles, one dataset
57tests passing in production
~90pages across the four portals

Outcome

The platform is live and in daily use by all four roles. We publish no adoption figures we can't stand behind, and we don't name the client — what we offer instead is a live walkthrough in the first conversation. It is the proof that Valtrin builds large, real, role-based systems, not only single-purpose tools.

A bright workspace with a management platform interface on screen
Illustrative — the live system is demonstrated in conversation, not screenshotted publicly

The standard

Nothing on this page is rounded up.

No logos of companies we haven't worked for. No testimonials we didn't earn. No metrics we can't demonstrate in a conversation. When a claim on this page can be shown, we show it — and when it can't, it isn't here.

How an engagement runs →

The next case study on this page could be your workflow.

Discuss a Workflow →