Overview
An engineering office does not sell anything you can put on a shelf. What it sells is Tuesday afternoon: a structural engineer’s four hours over a foundation drawing, a draughtsman’s day redlining a floor plan, a partner’s hour in a client meeting. The stock is people, the cost is salaries, and the two questions that keep a practice awake are who worked on what and when can we invoice it.
Most firms answer those from a spreadsheet somebody maintains in the evening. Nama has a module for exactly this shape of business — architecture and engineering offices, consultancies, audit and legal practices, design studios — and it follows one job from the offer that won it down to the hours spent on it, and back up into an invoice the client can be shown.
The phase-and-discipline grid
A firm describes itself once. The professions it employs are disciplines — architectural, structural, electrical, mechanical. The phases it delivers and bills in are milestones — concept, schematic, detailed design, tender documents. Those two multiply, and the multiplication is the heart of the module: four milestones across three disciplines is twelve work packages, each with its own estimated hours and its own estimated cost.
A firm that keeps winning similar jobs stores that grid once, and a new project inherits the whole thing in one action instead of being typed cell by cell. This is the structure the rest of the module hangs off.
From the quotation to the invoice
A project quotation prices the work: the priced lines, the tasks the firm expects to perform with the hours and rates behind them, the expenses it expects to incur, and a payment schedule. It commits nothing to the ledger — it is an offer. When the client accepts, one action turns it into a live project with its task list already populated.
Tasks hang off the project, each with a type, a milestone, a discipline and an executers grid — one row per employee, carrying that person’s planned hours, hourly rate and planned cost. That grid is where a project’s labour estimate actually lives.
Staff book their time on a daily sheet with start and stop buttons, one line per stretch of work. A manager then raises an approval, collects all the time waiting on projects they own, accepts or trims each line, and commits. That is the moment recorded hours become actual hours, the task’s cost is rebuilt, and the project’s totals refresh.
Out-of-pocket costs travel their own road: the employee claims the taxi and the printing against a task, and an accountant collects those claims into a document that books the cost and stamps them so they cannot be gathered twice.
The project invoice is where the design pays off. Rather than asking somebody to remember what has been billed, it carries collect buttons that sweep unbilled timesheet lines and unbilled expense lines straight into invoice lines. Nothing is re-typed and nothing is billed twice.
An hour is priced twice, on purpose
This is the part worth understanding before a firm commits, because it is where “the cost looks wrong” calls come from.
Cost comes from payroll. A timesheet line is valued at the employee’s constant monthly salary divided by the firm’s standard monthly hours — a 10,000 salary against a 200-hour month values an hour at 50 — and that is the figure that reaches the general ledger.
Revenue comes from the rate on the task. The hourly rate typed against the employee on the executers grid — a charge-out rate, a blended departmental rate, whatever the firm sells at — is what the invoice bills, multiplied by a markup factor set once in the module’s settings.
The two numbers travel different routes and are expected to differ. One is what the person costs you; the other is what you sell them for. The gap is the margin, and it is visible precisely because neither was derived from the other.
What it does not do
Three limits are worth stating plainly, because each is a reasonable thing to expect and none of them is there.
Nothing enforces a budget. The quotation prices the job and the project carries a contract value and estimated hours per work package — but no document blocks an overrun and no validation compares estimate against actual. The estimates are informational.
The scheduling is a record, not a plan. Project stages record targeted against expected periods and count late days. There is no critical path, no dependency between tasks, no resource levelling and no Gantt chart. A firm that needs true schedule management runs it elsewhere and records the outcome here.
Reporting is two timesheet reports. Employee TimeSheet and its detailed sibling, and nothing else — no shipped profitability report and no dashboard. The figures are all in the database, so profitability is a report you build or a question you put to the AI assistant, not a menu item you will find.
Construction is a different problem and has a different module: if the conversation involves a bill of quantities, a subcontractor or a payment certificate, the answer is contracting, not this.
The project screen, the matrix, the timesheet chain and the invoice’s collect buttons are documented screen by screen in the Project Management guide, including where project cost and revenue come from.
