Your operation's most expensive decisions no longer live in a spreadsheet.
Cromax connects your data, validates it and runs mathematical optimization models that deliver ready-to-execute plans. Compare scenarios in minutes and go with the best one, not just the only one you had time to build.
Born and proven in fresh fruit with Garces Fruit, Ranco Cherries and Prize Superfruits.
The problem
Margin-moving decisions, made against the clock.
Shift
Which lot runs on which line, and for which order, is decided by hand every shift, while raw material arrives different from what was expected.
Week
The weekly schedule gets redone on Tuesday, once reality has already broken it, and nobody remembers why that line was opened.
Purchasing
How much to buy and from which supplier is decided using last year's average and a single run.
People
Shifts for hundreds of people are built on a Sunday night in Excel, checking labor law compliance by hand.
How it works
One way to decide, from the shift to the season.
1
Data
Comes in as it is: ERP, sorter, databases or spreadsheets.
2
Parameters
Your operation's capacities, compatibilities and priorities, no code required.
3
Engine
Exact mathematical optimization: complex problems, answers in minutes.
4
Comparison
Named versions, compared side by side on their KPIs.
5
Decision
The plan goes back to the ERP and spreadsheets the data came from.
What you decide feeds back as data for the next run.
"What if the fruit comes in smaller?" gets answered here, before it does.
The platform
Simple up front. Serious under the hood.
Built for planners, not analysts: load, run, compare, decide. Under the hood, an optimization engine and industrial-grade data controls.
No black box
Full traceability: what data came in, what ran and why the result is what it is.
Simple interface
Load, run, compare and decide from a single screen, without depending on an analyst.
Robust
Multi-plant, multi-company and multi-user. Every run is logged and reproducible.
Adaptable
Your specific rule, the variety that can't go to that market or the line that only takes one size, goes into the model, not onto a waiting list.
Integration without an IT project
Connect databases and data warehouses, or upload files. Every load is validated before it reaches production.
Validation you can understand
Every data issue is reported with the exact row, the cause and how to fix it. Bad data stops the run instead of producing a wrong plan.
Feasibility preview
Before optimizing, see how much demand you can meet with current data and why the rest is lost, rule by rule.
Scenarios and versions
Every run is a named version you can compare by KPIs, export or use as the baseline for the next one.
Control center
One screen per decision: pick the plant, load data, tune parameters, run, read the KPIs and export.
Your decisions rule
Enable, force, prioritize or block markets by lot, order or line. Your overrides survive the next data load.
Incremental replanning
Start from the plan already issued and plan only what's new, without redoing what's already underway.
Back into your systems
The chosen plan is written to your database or exported to Excel or CSV. No more re-keying.
Security by company, plant and role
Data isolated per company at the database level, with roles and access profiles by module and plant.
Solutions
Three verticals. One way to decide.
Fresh fruit
The entire packing chain, from plant to shift.
Five chained modules: plant allocation feeds the weekly schedule, the schedule cascades down to each shift, the shift to each line's outlets and, at close, reconciliation aligns output with what the sorter measured.
Cherries
Blueberries
Stone fruit
Kiwi
Grapes
Avocados
Citrus
Apples
Pears
1
Plant allocation
Which fruit and which order goes to each plant?
Allocates fruit and orders to plants over a two-week horizon, prioritizing by shipping date, completing order lines and keeping grower groups together.
Order kilos fulfilled
Complete order lines
Unassigned fruit
2
Weekly planning
Which lines should run each shift of the week?
Decides which lines to open each shift and when to process each order, based on the harvest estimate and each line's real capacity.
Line-shifts opened
Planned boxes
Kilos fulfilled
3
Shift planning
Which lot goes to which line, in what order and for which order?
Assigns lots to lines and boxes to orders for one or two shifts, respecting capacities, the packing matrix, lot age and market eligibility. Minimizes variety and kit changeovers.
Demand fulfilled
Line utilization
Unplaced kilos
4
Outlet planning
Which size does each outlet handle, and with which kit?
For each line's lots, defines which size and color each outlet handles, when to change kits and the resulting line speed. It's the operator's run sheet.
Planned vs. shift hours
Kit changes
Idle outlets
5
Process reconciliation
How do you reconcile output with what the sorter measured?
At the end of the shift, moves boxes between compatible processes so each process's export % and size curve match the sorter, while keeping the mass balance.
Processes reconciled
Deviation before and after
Boxes moved
Market compliance built in
Phytosanitary and commercial requirements live inside the model: protocols for China, Korea and Vietnam, GGN, multi-residue, quarantine zones and lot-level conditions. A misspelled country won't switch off a rule.
-50%
in process reconciliation costs
-50%
in the cost of planning each shift
+1,000 h
freed up per year for the planning team
«We didn't just cut hundreds of hours of operational workload: we transformed the dynamics and mindset of our teams.»
Coming modules
Harvest estimation
Size and quality curves
Grower bin management
Box packing
Container planning
Supply
What to buy, from whom and when. With cost and service level in plain sight.
Plan purchases for your entire catalog month by month. Cromax decides orders by supplier while respecting lot sizes, pallets and containers, target coverage and your sourcing rules, and lets you compare scenarios before committing a single dollar.
What the model considers
Two-level lot sizing: units, pallets and containers
One supplier per container and single-supplier rules
Target coverage in months of safety stock
Lost sales or backorders, depending on your business
No avoidable stockouts
Efficient frontier between service level and cost
What it delivers
Purchase orders by month and supplier, with lots, boxes, pallets and arrival date
Inventory plan by material and period
Where every box will go, before optimizing
Service level
Stockouts
Container fill
Plan by material: month-by-month closing inventory, stockouts and purchase arrivals, with each material's balance and orders.
Scenarios side by side
What if container prices go up?
Every question is a version compared against the baseline plan. In the demo, making containers more expensive leads the model to buy with 15 containers instead of 18 and fill them more, in exchange for a slightly lower service level. The trade-off is in plain sight, with numbers.
18 → 15
containers
71% → 84%
average fill
100% → 99.4%
service level
Version comparison: baseline plan, customers who wait, strict coverage and expensive container, by their KPIs. Figures from the demo environment.
Shifts and staffing
The right headcount, on the right shift, within the law.
Builds each person's monthly shift schedule in 30-minute blocks and calculates how much staff you need of each contract type, at the lowest labor cost.
Shift planning
Assigns each person their shifts and breaks in compliance with Chilean labor law: maximum consecutive days, two Sundays off per month, contract hours and fair distribution of Sundays.
Staffing
Calculates how many people of each contract type each store needs to cover demand, at the lowest possible labor cost.
Coverage gaps
Labor cost
Contracted hours
+7%
in sales at stores with optimized shifts
-30%
in labor costs, with better demand coverage
Your next vertical
Your decision isn't on the list? We'll add it in weeks.
Cromax is a platform, not a set of closed modules. Data integration, validation, versioning, the control center and export are already built. For a new problem we only model your decision, so a new vertical arrives in weeks, not a year-long project.
Cromax
