AI for Operations · Solutions

Compare the work likely to arrive with the people, machines and space you have.

AI Demand and Capacity Forecasting projects the orders, jobs or visits expected over coming weeks and sets them against the hours, machines, vehicles or beds you can offer. A planner sees the tight weeks. Slack weeks appear too, each with a range. The numbers are estimates for a person to adjust.

Share your requirements

What it is

An operations head usually carries two guesses in mind. One is how much work is coming. The other is how much the shop, the fleet or the ward can handle. When the two disagree, overtime follows in March. Idle machines follow in May.

This capability writes both guesses down and compares them. It projects demand from your order and job history, lists the capacity of each resource, and shows the gap by week. The planner then adjusts for what no software can know, such as a festival, a tender or a customer who has gone quiet.

Where it goes wrong today

The problems this is meant to ease.

Overtime one month and idle hands the next

Work arrives in waves that nobody wrote down. The shop pays overtime in the rush and carries unused hours in the lull, because staffing was set by feel.

Promises made without a look at capacity

Sales confirms a date for a rush order. Only afterwards does the planner find that the press is already full for three weeks.

Seasonality that surprises every year

The peak comes back each year, yet it is always treated as news. Hiring and stocking start late. Customers feel the delay.

Capacity counted from the brochure

A machine is listed at full hours, ignoring service days, setup and breakdowns. The plan looks fine on paper. It fails on the floor.

Equipment bought or hired on instinct

A second vehicle or machine is purchased after one bad month. No record shows whether the shortage was a pattern or a single event.

Inside the module

What AI demand and capacity forecasting does.

Demand forecast from your history

History projects the weeks ahead. Repeat customers, open quotes and the usual swings of the year are shown separately, so each source stays visible.

Workload predicted from patterns

Order counts are only the start. Expected hours, trips or tests are estimated from past sizes and durations, so a large job counts for more than a small one. A planner sees each job type’s usual workload.

Seasonal and peak demand

Recurring peaks, such as the weeks before a holiday or the end of a financial period, are marked on the calendar. Flagged one-off spikes stay out of the pattern.

Staff, equipment and space requirements

Each week gets estimates for machine hours, technician days, trips, bays and beds, with planned maintenance and approved leave subtracted so totals match what you can really offer.

Demand against capacity by week

A simple view marks weeks where projected work exceeds what you can do, and weeks where capacity sits empty. Split it by site, line or team.

Ranges and planner overrides

Each projection shows a low and high case. Is a large customer pausing orders? Enter it. It is marked manual. The model’s figure stays beside it.

What-if scenarios and shortage alerts

Add a second shift, rent a machine for a month or hire two temporary hands, and see which weeks close. Alerts name the weeks heading into shortage. Nothing is committed.

Where it gets used

What this looks like in a working week.

Truck trips at a logistics firm

At the logistics firm, expected consignments by week are compared with available vehicles and drivers, and a tight fortnight shows early enough to arrange a hired truck.

Bed and theatre load at a hospital

Admissions and surgeries meet beds and theatres. The administrator spots weeks for deferring elective cases.

Machine hours at a workshop

Open quotes and repeat orders are projected against lathe and press hours, and the owner decides whether to add a shift or turn down a rush job.

Wash load at a commercial laundry

Linen volume from hotels and clinics follows seasons. The plant manager sees which weeks need extra staff and plans leave around them.

Field visits at a service company

Expected service calls by area are compared with technician days. Routine visits move out of tight weeks.

Who touches it

The people who open this every day.

Operations or planning manager

Reviews the weekly gap and decides on overtime, subcontracting or a deferred delivery date, then records the reasons for the next planner. The reason stays on file.

Branch or workshop head

Checks the month ahead before a big booking. A quick check saves a broken promise.

Sales head

Checks slack weeks before promising a delivery date. Checks tight weeks before accepting a rush order.

Owner or finance lead

Uses the projection to time hiring, equipment purchase or a rented shift, with the range visible so nobody mistakes a forecast for a promise made to customers. Nobody sees a false certainty.

How it is set up

From first conversation to daily use.

Decide what is being forecast
Name the demand to study, such as orders, jobs or visits, and the capacity to compare with, such as machine hours, vehicles or beds. Agree the weekly unit with the planner.
Gather and clean the history
Collect order, job or booking records with their dates and sizes. Fix duplicates and missing dates, and mark known one-off events, so the pattern is not distorted.
List capacity honestly
Record each resource with maintenance, leave, holidays and the extra hours you can usually call on. Each figure needs planner confirmation.
Run the projection on a past period
Project a past period. Compare it with what really arrived. Show the gaps to the planner and the owner, and talk through the reasons.
Go live with the planner in charge
The weekly gap view starts alongside your current planning method. Adjustments are made by the planner with reasons recorded, and the plan used for decisions carries the planner’s sign-off, never the software’s.
Review accuracy every quarter
Compare each past projection with actual demand and capacity used. Tighten the inputs that caused large misses, and note which kinds of work remain hard to predict.

Measuring it

How you would know it is working.

Record where you stand before it goes live, so the change is measured against your own figures.

Forecast accuracy is the weekly gap between projected and actual orders or jobs, compared with how far your planner’s earlier estimates missed the actual records.
Capacity utilisation is hours or units used set against hours or units available, and last year’s attendance, machine and booking records supply the earlier figure.
Service delays are jobs finished later than the date promised, counted and compared with your delivery or completion records from before go-live.
Weeks of overtime or hired capacity, counted from payroll and purchase records for the same period.
Shortage alerts that proved correct, reviewed with the planner after each quarter.

Used responsibly

Where a person stays in charge.

Numbers belong to the planner

The projection is a draft. The planner questions, adjusts and signs off. Hiring, purchase, subcontracting and customer commitments are never made by the software.

Estimates, not promises

Every projection carries a range. Wide ranges are flagged as a warning. The model sees only past patterns, and it hears no unrecorded conversations or news.

Capacity data must be current

Stale lists make the gap view wrong. The record owner is reminded once a list sits untouched past a set period, so gaps are caught before they distort the totals.

Staffing limits belong to your adviser

Working hours and rest rules vary between countries and trades, and only your own legal adviser can say what applies to your staff. Figures only, no employment or legal advice.

How it fits

It is one part of a system, not an island.

A module earns its place by what it passes to the next one. These are the connections we set up most often.

Production planning
Order management
Sales forecasting
Technician scheduling
Attendance and time tracking
Your ERP or booking software

Already built in

Products that include it.

Built for you

How we would put it in.

Common questions

Questions owners ask about AI demand and capacity forecasting.

How much history does the projection need?

Two full seasonal cycles give a fair basis. That means two years for a yearly pattern. Six months of records can still be used. Expect wide ranges and treat them so.

Can it predict the loss or gain of a big customer?

No. Enter a large account by hand instead. The model sees only past patterns.

Who owns the numbers?

Your planner does. The projection is a draft, not a decision. The software hires, buys and promises nothing.

How accurate will it be?

Accuracy depends on steady work and clean records. Accuracy is measured against your own past weeks and shown to you, so you can decide how much weight to give it.

Which records must we hand over?

Order, job or booking history with dates and sizes, a list of resources with their available hours, and planned leave and maintenance. Missing data shows as a gap. It is never guessed.

What does it connect to?

Your ERP, booking, attendance or production planning software, through whatever each one opens up. File exports work too.

Is it any use if capacity is flexible, like freelancers?

Yes, with a record of extra capacity. Note how much extra you can usually call on, and how soon, because without that the gap view overstates the shortage and your planner will rightly ignore it.

When is a spreadsheet enough?

With one site, one product and steady orders, a spreadsheet and your own judgement do well. The software starts to help with several sites, mixed products or seasonal swings that surprise you every year.

Who needs it most

Businesses that ask for this.

Next to it

Modules that work with it.

Talk to us

Tell us how you handle AI demand and capacity forecasting today.

A spreadsheet, a register or another system: say which, and we will tell you plainly what is worth changing.

No obligation
A reply within one business day
Your details stay private

Start with your requirements

Seven short steps that tell us what you need. An engineer reads them before getting in touch.