What you should do next
Scheduling and optimization for small and mid-size businesses: the best choice inside the rules you work under, from machine hours to the delivery dates you've promised.
Our AI assistant asks about your project and can build you a free plan.
Can you beat our route?
One truck, eight stops. Tap the stops in any order to plan the route, then see how it compares with the optimized one.
A street map with a depot and eight stops. Choose the stops one by one to build a route; the truck returns to the depot at the end. A counter shows the miles and the fuel cost as you go.
How to read the map
- Depot: the truck starts and ends here
- A stop to visit
- A stop you chose, numbered in your order
- Your route
- Optimized route
- Streets: one block is one mile
Optimized route: 36 miles · fuel $18
Example numbers, not a promise Against taking the stops in the order listed (40 miles): 4 miles saved a trip × 250 weekdays × $0.50 of fuel a mile ≈ $500 a year.
Your business has more rules than this game. Let's talk.
Illustrative dataThe problem
A forecast tells you what's coming. Optimization tells you what to do about it, and that's where the savings are: which jobs go first, who works where, which route each truck takes, and who gets served first when you can't serve everyone.
Usually one experienced person makes these calls every day, with a spreadsheet and good judgment. That works until there are too many moving parts to keep in one head, often the day a new rule appears. The schedule still looks fine on the floor, but it starts costing the business more, and nobody can see why.
How we approach it
Before any math, we write down three things:
- What you want more or less of: profit, time, cost, late deliveries. Pick one thing, or a clear trade-off between two. "Everything at once" isn't a goal.
- The rules you can't break: machine hours, people, shifts, how much a truck holds, contracts. That includes the unwritten ones, like the customer who won't take a delivery in two parts.
- What you're free to change: the decisions that are yours to make.
Much of the value shows up in that conversation, before any model exists. Often a rule everyone treats as fixed turns out to be a habit, and dropping it saves more than any clever math.
Then we build a tool that checks the options properly and recommends one, with the reason it beat the others. If you can't question a recommendation, nobody will follow it, and they'd be right not to. The person running the schedule knows things the model doesn't, and has to be able to overrule it with eyes open.
When there are too many options to check one by one, machine learning helps narrow them down, for example by learning how long each job takes in practice, since nobody believes the standard times. The optimizer still makes the choice.
Most of the speed in optimization comes from being honest about constraints. We separate the hard ones, which can never be broken, from the soft ones, which only cost something to bend. That reaches a better answer sooner than searching every possibility.
A worked example
Three crews, seven jobs and one week, with a few rules nobody gets to ignore. As in real optimization work, the best answer always has to fit inside your rules.
The clinic and the store only let crews in on Tuesdays. The school is only empty on Mondays and Fridays. Nobody works Saturdays. Each crew drives farther to some jobs than to others, and the drive is time you pay for. Build the week yourself, then press Optimize and compare.
Build the week
Three crews, seven jobs. Drag each job into a day, or tap a job and then tap a slot. Then press Optimize and compare.
A weekly plan with three crews as rows and Monday to Saturday as columns. Pick a job from the list, then pick a crew and a day for it. Some jobs only fit on certain days. A counter shows the total hours on the road.
Showing: optimized week
Saturday: nobody wants this shift. A real rule.
A crew drives 1 hour to a job on its own side of town, and 3 or 4 hours to the others.
Jobs to place
Optimized week: 9 hours on the road
Example numbers, not a promise Against putting each job on the first free day (20 hours): 11 hours saved on the road a week × 50 weeks = 550 hours a year; at $40 an hour for a crew, about $22,000.
Your business has more rules than this game. Let's talk.
Illustrative dataFixed scope
Optimization Starter
What you get
- One routing or scheduling decision modeled with your real constraints: crews, time windows, distances
- A better plan for a sample week, side by side with the one you use today
- The driving time and fuel cost of both plans on that same week
- A verdict on what a full solution would take, or whether a simpler fix is enough
- The written method and the model, so the result can be checked
What we need from you
- The one decision you want to make better, and time with the person who makes it today
- A sample week of real work: the jobs, where they are, their time windows, and the crews or vehicles
- The rules that can never be broken, and the ones that only cost something to bend
What is out of scope
- Running the new plan automatically every day
- More than one decision
- Live GPS or real-time dispatch
- Connecting to your scheduling software
- Ongoing changes after handover, which is what the care plan is for
- How long it takes
- Three to four weeks. Start dates depend on current availability.
- What happens after
- Thirty days of fixes are included, then the optional care plan. If the new plan earns its keep and you want it running every week, that is a bigger project, which we scope and quote separately.
- Price
- We give you a fixed price before we start, usually within two business days.
Bigger than one decision
The Optimization Starter is a fixed-scope first step: one routing or scheduling decision, modeled with your real constraints, and a better plan you can compare with today's.
Bigger optimization work varies a lot. The same request, "help us schedule better", can be a two-week problem or a six-month one, depending on how your constraints interact. It might mean several decisions at once, a plan that re-runs every day, or a model wired into your scheduling software. We scope that work properly and quote it per project, once we know what we're looking at.
Start here: Tell us the decision you are trying to make.
We don't need your data or your software yet, just the decision and what makes it hard. The conversation is free and short, and it usually shows whether this is a real optimization problem or something simpler.
How it usually attaches to a package
- After a Forecast Starter You know what's coming, and now you decide what to do about it. It's the natural next step.
- Alongside a Dashboard Sprint The recommendation has to be visible where the decision gets made, or it does not get used.
- After data engineering If the constraints live in three systems that disagree, that gets fixed first.
Need something bigger?
A model that re-plans your schedule every week is a systems project in its own right, and we scope it carefully.
If the decision has to re-run at 6am every day with someone on call, tell us and we'll tell you whether we're a good fit.
Every project is handled directly by a member of our team.
If you let us publish what we build together, you get a discount. The work has to be something we can publish, and you approve every word before it goes live. Our team applies the discount when we quote.
Questions people ask
Is this the same as a forecast?
No. A forecast predicts what will happen; optimization chooses what to do. They work well together: forecast the demand, then decide how to meet it. A forecast on its own leaves the hard part to you.
Do I need perfect data?
No, but you do need to know your real rules. A missing rule hurts far more than slightly messy numbers: a model that doesn't know about a contract will happily recommend breaking it.
How do I know the answer is right?
You'll see why it beat the alternatives, and you can change an assumption and watch what moves. A recommendation that can't be explained shouldn't be trusted, so we build ours to be explained.
What if the recommendation is unworkable in practice?
Then a rule is missing and we add it. Expect that in the first week: it's a normal part of the process.
Is there a fixed-scope way to start?
Yes: the Optimization Starter, for one routing or scheduling decision. Anything bigger can hide very different problems, so we scope it first and then quote a fixed price.
Which decision is hardest to get right today?
Tell us the decision you're trying to make. If we're not the right people, we'll say so.
Our AI assistant asks about your project and can build you a free plan.