How it works
Bring a week of your loads. Get schedules that sharpen every cycle.
A guided setup, no six-month implementation, no data team required. You bring the loads and set the rules; each training cycle tries thousands of schedule variations on your real loads and keeps what scores best. Here is exactly what your dispatchers will do, step by step.
First
Your loads: bring a week, or enter a day by hand.
Drop in a week of recent loads and drivers - origin, destination, appointment window, pay, equipment - and you have a dispatch network ready to train on. No field-mapping workshop: you point each column at its field in one sitting, and anything you leave unmapped is left out. No week of history handy? Enter a day by hand and train on that.
The point: you do not need a clean data warehouse. A week of operational reality is enough to train a useful brain. Your dispatchers can be in the app the same day.
There is no professional-services phase: the upload, the mapping, and your first training run all happen in the same sitting, inside the product.
Then
Your rules: define a good day, your way.
Pick from the templates - revenue-first, on-time-first, balanced - or compose your own. Tick the rules that count for you (on-time bonus, empty-miles penalty, driver-home-time bonus, customer-priority multiplier), set the weights, save. Your idea of a good day, not a vendor's, is what every schedule trains against.
Need a rule that does not exist? Add it in the editor and the app checks it before training. Bad rules never reach a training run.
Every weight is yours to change: edit a rule, re-run training, and compare the result the same day. No change order, no vendor ticket, no contract amendment to move a number.
Next
Let it learn: your schedules sharpen against your rules.
Hit Start. You watch the score climb in real time and the leaderboard reshuffle as each cycle tries different scheduling moves, keeps what worked, and drops what did not. Stop early or let it run overnight - results are saved every cycle.
Training runs finish in minutes to hours depending on fleet size and settings. Each cycle builds on the champions of the last, so the score keeps climbing while the run is live.
The score is computed against your rules on your loads, not someone else's benchmark. When it climbs, that is your schedule getting better by your own definition of better.
Finally
Go live: promote a brain and ship decisions.
Promote a brain to Live. From that moment new loads arriving in your TMS get scored in milliseconds. Dispatchers see the brain's recommendation alongside its reason - they can accept, override, or request a replay. Every decision is logged and replayable, end to end.
No spreadsheets. No manual juggling. The dispatcher's job shifts from eight-tab triangulation to the exceptions that actually need a human.
Every decision can flow straight back into your TMS or reporting tools as it happens - assignments, overrides, and scores alike. Ask us for the current integration reference for your IT team.
Your people stay in charge of the board: an override is recorded and pinned, and the next recompute respects the human call instead of fighting it. NEO advises; your dispatchers decide.
Training & change management
Your dispatchers are productive in an afternoon.
Onboarding is designed as 2 to 3 hours of guided setup per dispatcher, plus one week of side-by-side use alongside their existing process. There is no separate certification, no week-long training program, and no new tool to learn end-to-end - the dispatcher's job is still "review and approve a schedule." NEO just removes the spreadsheet step.
Recommended change management: weekly review of the brain's recommendations and any dispatcher overrides for the first two weeks. Overrides are recorded and pinned - the next recompute respects the human call - and the review is where you spot the scoring rule or weight worth adjusting before your next training run.
Total adoption cost in dispatcher hours: a half-day per dispatcher in week one, then steady-state. No multi-week training sprint.
Integrations
TMS-agnostic by design.
NEO connects to any TMS that can POST JSON over REST or receive an outbound webhook - including McLeod, TMW, Trimble, and homegrown systems. Specific TMS adapters are added during onboarding to match your install. If your TMS only speaks EDI or CSV exports, our import wizard handles the seam.
See your fleet's brain in 10 minutes.
Bring a week of recent loads, or enter a day by hand. We will show you a trained brain against your rules before you leave the call.