Peaks & ValleysConsulting
Home/Services/Modelling

Modelling and simulation

Some decisions are too big for a spreadsheet and too specific to buy software for. A capacity commitment. A fleet replacement cycle. A pricing structure. A plan that locks in ten years of cost.

We build the model that answers it, on your data and your rules. Our technology founder has built modelling and simulation software for regulated field operations, and works in ArcGIS Pro where the decision runs over geography instead of over rows.

01 / The problem

The biggest decisions get the worst tools.

There's an odd pattern in most operations. Routine work gets proper software. The decisions with the most money attached get a spreadsheet someone built once, which no one fully understands and no one dares change.

It happens because those decisions are specific. No vendor sells a model of your operation, so the estimate gets made by hand, the assumptions live in one person's head, and running a second scenario takes a week. So only one scenario ever gets run, and it's usually the one someone already believed.

A model earns its cost the first time it lets you ask what if and get an answer the same afternoon. That's the entire point of building one.

02 / What we do

Everything this covers.

Scenario and what if analysis

  • Run the same decision ten ways and compare the outcomes side by side
  • Assumptions exposed as inputs you can change, not buried inside a formula
  • Sensitivity analysis, so you know which assumption drives the answer
  • Scenarios saved and rerun next year against what really happened
  • A record of which scenario was chosen, and on what basis

Forecasting and capacity

  • Demand, volume and capacity forecasting from your own history
  • Staff, fleet, equipment and crew sizing
  • Seasonal and cyclical patterns modelled instead of averaged away
  • Forecasts with their error stated, because a forecast without one is a guess
  • Capacity commitments tested before they're signed

Pricing and cost models

  • Unit economics modelled instead of averaged across the book
  • Cost to serve, by client, segment, job or region
  • Pricing structures tested against your real mix before you launch them
  • Margin modelled across a portfolio instead of invoice by invoice
  • Break even and payback on a capital decision

Optimisation and scheduling

  • Scheduling and sequencing under real constraints
  • Routing, allocation and assignment problems
  • Cost and margin optimisation across a portfolio of jobs, sites or clients
  • Constraints that reflect regulation, contracts and physical reality, not just arithmetic

Spatial models

  • Models that run over geography instead of over rows
  • Terrain, access, distance and haul cost built into the answer
  • Territory, coverage and catchment modelled instead of drawn by hand
  • Results delivered as maps and as data, in ArcGIS Pro
  • Jurisdiction and regulatory boundaries applied inside the model, not checked afterwards

Industry specific models

  • Models built for the actual rules of your sector, not adapted from a template
  • Multi site operations where every location carries its own constraints
  • Field and fleet operations where distance and terrain drive the cost
  • Regulated industries where the binding constraint is legal instead of physical
  • Whatever your operation is, modelled the way it actually runs

Built to be trusted

  • The method documented, so a result can be defended to a regulator, a board or a client
  • Inputs, assumptions and version recorded with every run
  • Validated against outcomes you already know, before anyone relies on it
  • Handed over readable, so your team can change an assumption without calling us
Python · Simulation and optimisation · ArcGIS Pro · AI applied to operational systems
03 / How it works

The order we do it in.

01

Find the real decision

Not what model do you want. What decision are you making badly, how often, and what does getting it wrong cost. Plenty of modelling projects should be a two page analysis instead, and we'll tell you when yours is one.

02

Model the rules, not the outcome

The constraints, the costs and the regulation as they are. A model built to produce the answer someone already expected is worth nothing and worse than nothing.

03

Validate against the known

Run it against periods and outcomes you already have. If it can't reproduce last year, it won't predict next year, and we'll say so instead of ship it with a disclaimer.

04

Make it rerunnable

Something your team runs themselves, as often as they like, with the assumptions on the screen instead of buried three sheets deep.

04 / Fit

Who this is for.

A good fit

  • Companies making capital decisions on a spreadsheet no one fully understands
  • Businesses that can only ever afford to run one scenario
  • Anyone whose forecast is last year plus a percentage
  • Operations planning work across terrain or territory instead of across a list
  • Companies that must defend an estimate to a regulator, a board or a lender
  • Field, fleet and multi site operations
  • Businesses whose pricing has never been tested against their real mix
05 / Questions

What people ask about this.

How is this different from a really good spreadsheet?
Often it isn't, and where a spreadsheet is the right answer we say so. The line gets crossed when the logic has to branch, when geography or terrain is part of the calculation, when you need fifty scenarios instead of three, or when the file has become something only one person can safely open. At that point the spreadsheet is a liability holding a decision that matters.
How do we know the model is right?
You don't take our word for it. Before anything is relied on, the model runs against periods where you already know what happened. If it can't reproduce outcomes you can verify, it doesn't get used. Every run records its inputs, assumptions and version, so a result can be reproduced and defended months later when someone asks where the number came from.
What kinds of businesses is this for?
Any operation where a decision is too complex for a spreadsheet and too specific to buy off the shelf. That covers capacity and fleet planning, pricing and cost to serve, scheduling under constraints, territory and coverage, and regulated field operations, where the decisions are spatial, constrained and long dated all at once. If your hardest decision gets made once a year by one person with a spreadsheet, it qualifies.
06 / Related

The work next door.

These almost always come up together. An audit that finds a process problem usually finds a data problem underneath it.

Start here

Start with the audit.

Every business is different, so we scope the audit that way. We spend time inside your operation and watch the work actually happen. You get a written map of every valley, what it costs you a year, and what it would take to flatten it. That document is yours whether or not we build anything.

Emailadaignault@peaksnvalleysconsulting.com
Phone(250) 423-9511
FoundersKamila Domanska · Alex Daignault
Based inNelson, British Columbia
ServingCanada and the United States