Far Retail · Demand Intelligence
Produce and replenish what customers will actually buy.
A forecast and a recommended quantity per site and day, with the reasons behind it.
Working with
- Store
- Gràcia
- Product
- Sourdough loaf, 500 g
- Plan
- Example plan for a Friday
Units per day: the last 14 days, then the Friday plan
- Units sold
- Forecast demand
- Expected range
- Recommended production
Expected range: 53 to 66 units. Demand will most likely fall within it.
Recommended production
64 units
The current plan is 80, so this means baking 16 fewer.
How 64 is calculated
- Typical Friday, last 8 weeks
- 63
- Recent trend, last 4 weeks
- −4
- Forecast demand
- 59
- Buffer so the shelf doesn't run out
- +5
- Recommended production
- 64
Forecast plus a small buffer, in case demand runs high.
Previous day (Thursday)
- Recommended
- 52
- Baked
- 64
- Sold
- 51
- Left over
- 13
Each day, the recommendation is compared with what was baked and sold.
- Check the forecast.
- See how the quantity is calculated.
- Compare it with what was sold.
Every day, every site, the same costly decision.
Too much means waste. Too little means lost sales.
Your current plantoday's quantity
Far Foundationhistory, stock, calendar, signals
Forecast and constraintsFar Decisions
Recommended quantityfor a planner to approve
The right model for each part of the job.
Forecast and optimize
Up to 14 state-of-the-art forecasting models, combined
- Statistical time-series methods, gradient-boosted trees and probabilistic forecasting
- Far weighs them and keeps the best of each
- Optimization under your constraints
Understand and explain
Language models do the reading
- Map your data and planning rules
- Explain each recommendation
Getting started.
- For
- Planners and operations teams in multi-site companies
- Inputs
- Sales per site, stock, waste, promotions, calendar, weather, events
- Who decides
- A planner reviews and approves
- Success
- Waste, availability and margin against your current plan, in euros
Let's look at how you plan production and replenishment today.
In a first conversation we look at the problem, the information you already have and whether there is a good first use case to measure.
or write to hello@farsystems.eu