Features
Platform capabilities
Orthantic is built on six core capabilities: Orthantic AI, the full planning cycle in one platform, statistical forecasting, optimisation, collaborative planning and a business-readable formula language. A full capability reference follows.
Core capability · AI
Orthantic AI: an agent for analysis and planning
An AI agent on every dashboard answers questions, explains results and prepares changes for a person to approve. AI-assisted model building reduces time to value.
Answers about any figure, with tables and charts.
The drivers behind any variance, explained in business terms.
The agent prepares a change. A person approves it, and it can be undone.
AI-assisted model building reduces time to value and accelerates your business planning.
Core capability · one platform
The full planning cycle in one platform
Planning, statistical forecasting, optimisation and analytics run on the same data, in the same model. There is no need to export data to a forecasting tool, re-import the results and export again for optimisation.
Forecast and optimisation results update the plan directly
When a forecast or optimisation run completes, its results are written into the planning model, and every figure in the model that depends on them, such as supply, inventory, revenue and cost, is recalculated. There is no waiting for results to arrive by email and no Excel export to carry them from one tool to the next.
Standalone tools
- Run forecast
- Export to Excel
- Email to planning
- Import into the plan
- Export for optimisation
- Email the results back
- Update the plan
With Orthantic
- Run forecast or optimisation
- Results written to the model
- Dependent plans update
Core capability · forecasting
Statistical forecasting, built into the plan
Generate demand and sales forecasts from your history using proven statistical methods, and use them directly in the planning model.
Why better forecasts matter
-
Right stock, right service
Reduce stock-outs and excess inventory by planning supply against a more accurate view of demand.
-
More reliable financial plans
Build revenue, cost and cash plans on a sound demand baseline.
-
Earlier warning
Identify trends and seasonal patterns before they affect service levels or the P&L.
-
Time for judgement
Planners spend their time reviewing and refining forecasts rather than producing them.
Why forecast in Orthantic
Forecasts run on the data already in the planning model, and the results are written straight back into it. Every figure in the model that depends on the forecast is recalculated, with no exports, emails or re-imports.
Several forecasting methods are run and compared on your own history. The best-performing method is selected for each series, improving accuracy across products and markets that behave differently.
Forecast at the level where demand patterns are clearest, such as category by customer, or group like products and forecast them together. Results are automatically disaggregated to the detailed level, such as SKU by store.
Planner overrides are held in their own layer, separate from the statistical baseline, so the accuracy of the baseline and of each adjustment can be measured over time.
How it works
- Historical data
- Choose level and horizon
- Run and compare methods
- Review and adjust
- Plan
Select the hierarchy level, horizon and frequency. Orthantic runs several statistical methods, scores each against your history using error metrics such as MAPE, RMSE and MAE, and presents the results side by side so planners can review the comparison before accepting a forecast.
Forecast at the level that fits your data
Data is often held at a detailed level, such as SKU by store, where individual series are too sparse or volatile to forecast reliably. Orthantic lets planners choose the level at which to forecast, independently of the level at which data is stored.
- Data at SKU × store
- Forecast at category × customer
- Automatic disaggregation
- Forecast at SKU × store
-
Forecast at a higher levelAggregate
With data held at SKU by store, forecast at category by customer, or any other combination of hierarchy levels, where demand patterns are more stable.
-
Group like productsCombine
Combine similar products, such as variants, pack sizes or low-volume lines, and forecast them as a single group.
-
Automatic disaggregationSpread
Forecasts made at a higher level or for a group are automatically disaggregated back to the detailed level, ready for planning.
-
Forecast at detailDetail
Where volumes support it, forecast directly at the level the data is held, such as SKU by store.
Statistical baseline, consensus plan and actual
Thousands of units. Actual stops after July; the prediction interval opens from there.
Illustrative figures. The prediction interval widens with the forecast horizon.
Core capability · optimisation
Linear optimisation, inside the planning model
Turn a plan into a decision. Orthantic finds the best allocation of any limited resource, such as production, supply, marketing budget, capital or people, against your objectives and within the constraints already defined in your plan.
Why optimise
-
Lower cost, higher margin
Identify the allocation that best meets your objective, rather than the first plan that works.
-
Feasible plans
Every result respects the limits in the model, such as capacity, budget, service levels and business rules.
-
Balanced trade-offs
Weighted goals balance cost, service and other objectives in a single optimisation.
-
Decisions that keep up with change
Re-optimise whenever conditions change, as part of the regular planning cycle.
Where you can use it
-
Supply chainProduction
Allocate production and sourcing across plants and suppliers at the lowest cost, within capacity and service levels.
-
Marketing mixMedia
Split the budget across TV, radio, print and digital to get the highest expected return, within the total budget and each channel's limits.
-
FinanceCapital
Choose which projects to fund and how much to give each, against a capital limit and the return each one is expected to bring.
-
WorkforcePeople
Plan headcount and shifts to cover the expected workload at the lowest cost, within hiring and working-time rules.
Why optimise in Orthantic
Capacity, demand, budgets, costs, expected returns and service levels are read directly from the planning model. The optimisation always works from the current plan, with nothing exported.
Update a figure such as factory capacity and run the optimisation again. There is no separate tool, no data movement and no downtime between the change and the new result.
Set the relative importance of objectives such as cost, service level and utilisation, and the optimisation balances them according to your priorities.
The recommended allocation is written into the model as a scenario to review and compare. Once adopted, every figure in the model that depends on it is recalculated, with no hand-off to another tool.
How it works
- Drivers, limits and costs in the model
- Set objectives and weights
- Run optimisation
- Recommended allocation
- Updated plan
- Financial impact
What you can model
-
ObjectivesGoal
Minimise cost, maximise margin or return on investment, or combine several objectives using weighted goals.
-
DecisionsVariables
Continuous quantities as well as whole-number and yes/no decisions, such as batch counts, minimum run sizes, or opening and closing a line.
-
ConstraintsLimits
Capacity, supply, demand, service levels, budgets and business rules, taken from the planning model.
Balancing competing costs
Setup cost falls and holding cost rises as batch size grows. Total cost is lowest where they balance.
Illustrative figures. The optimiser finds the point of lowest total cost, here a batch size of 4,000 units, while respecting every constraint in the model.
Core capability · collaboration
Collaborative planning in a shared model
Contributors across regions and countries plan in one shared model rather than exchanging files. Each person enters their own inputs directly, with role-based access controlling what they can see and edit.
Every manual change is recorded with the user and a timestamp, and versions and scenarios are held in the same model, so reviews work from one set of numbers.
Core capability · modelling
Business-readable formula language
Every calculated value in Orthantic is defined by a readable formula. The syntax uses
familiar spreadsheet concepts: arithmetic, comparisons, if and a library of
named functions that behave like their spreadsheet equivalents.
Example formulas
# Revenue, from two different modules
"Sales History"["Sales Value"] = "Sales History"["Quantity"] * "SKU Master"["Sales Price"]
# A driver-based cost line, with a floor
"Shipment Plan"["Freight Cost"] = max("Shipment Plan"["Shipped Units"] * "Freight Rates"["Rate per Unit"], "Freight Rates"["Minimum Charge"])
# Last month's closing becomes this month's opening
"Inventory Plan"["Opening Stock"] = lag("Inventory Plan"["Closing Stock"], 1, 0)
# Each SKU's share of its category
"Sales Plan"["Mix %"] = round("Sales Plan"["Units"] / aggregate("Sales Plan"["Units"], "SUM", over=[Category]), 4) * 100
Each input is qualified by its source module, for example
"SKU Master"["Sales Price"], so every formula identifies where its inputs
come from.
Business users can follow the logic behind any figure without specialist training. Rules are maintained by the model-owning team, so changes can be made without a hand-off to IT or a development cycle.
Business rules are defined as formulas rather than scripts or stored procedures. The formula is both the definition and the implementation, reducing reliance on custom code and on the individuals who wrote it.
Orthantic records which rules and inputs each calculated value comes from, so the model always knows what depends on what. Every manually entered value records the user and a timestamp, supporting audit and review.
Function library
More than forty-five functions covering arithmetic and logic, time intelligence, cross-module lookups, aggregation to any level of any hierarchy, and text handling. New functions are added with each release and are available to existing models without rework.
| Group | Function | What it does |
|---|---|---|
| Arithmetic & logic | max |
Returns the larger of its arguments, for example to apply a minimum to a calculated cost |
| Time intelligence | lag |
Returns a value from an earlier period, for example to roll closing stock into the next opening balance |
| Time intelligence | rollingtotal |
Runs a total or average across a moving window of periods |
| Lookup & aggregation | aggregate |
Aggregates a value to any level of any hierarchy, for example to calculate share of category |
| Rounding & maths | round |
Rounds to a stated number of decimal places |
Capability reference
| Capability | Description |
|---|---|
| Orthantic AI | |
| AI agent on dashboards | Answers questions about the plan in natural language, with tables and charts |
| Root-cause analysis | Explains why a figure moved by following the model's dependencies to its drivers |
| Changes with approval | The agent prepares changes to input data; a planner applies them, and each one can be undone |
| Dashboard actions | The agent offers the dashboard's own actions, such as running a forecast, for the planner to run |
| AI Model Builder | AI-assisted model building, for a faster time to value |
| AI conversation log | Every conversation, step and proposed change is available to administrators for review |
| Modelling | |
| Multidimensional modelling | Plan across products, customers, locations, accounts and time, at the level of detail each process requires |
| Hierarchies and roll-ups | Plan top-down and spread to detail, or plan at detail and consolidate, within the same model |
| Formula language | Business rules written once in a readable, spreadsheet-style syntax and applied across the model |
| Function library | More than forty-five functions covering arithmetic, logic, time intelligence, lookups, aggregation and text, extended with each release |
| Cross-module references | Reference values from other modules by name, so every formula shows the source of its inputs |
| Time intelligence | Rolling periods, year-to-date, prior-year comparisons and period roll-forward |
| Planning | |
| Grid entry and import | Enter values directly, paste from a spreadsheet or import a file, all handled the same way |
| Scenario planning | Build and compare alternative scenarios side by side, retaining those not selected |
| Versioning | Snapshot plans as the cycle rolls forward and compare against prior versions |
| Forecasting and optimisation | |
| Statistical forecasting | Generate a baseline forecast from history, with outlier treatment applied before fitting and best-fit method selection per series |
| Hierarchical forecasting | Forecast at any level of a hierarchy or for groups of like products, with the result automatically disaggregated to detail |
| Linear optimisation | Linear and integer optimisation with weighted goals, using constraints taken from the planning model: production, sourcing, marketing mix, capital and workforce |
| Analytics and reporting | |
| Dashboards and reports | Role-specific views built from the same underlying planning data |
| Governance and security | |
| Cell-level audit trail | Previous value, new value, user and timestamp for every manually entered figure |
| Role-based access control | Control access by role, down to the data each user is permitted to view and edit |