AI-Informed Demand Forecasting

At Moselle we provide various AI-informed forecasting models which can automatically plan your inventory for up to 12 months.

These models are excellent starting points for your business and intelligently learn from your historical sales data. As you continue to use Moselle, the forecasts will improve over time.

While there are many factors to consider, the best indicator when choosing a model is your MAPE score. Once you apply the model, you’ll be able to evaluate this.

Different forecasting methods are utilized by model. If you’d like to learn more about Bottom-up or Top Down methods read through our guides.

Learn more about our models below:

Moselle Model Name	Rome	Athens	Bombay	Venice	Hong Kong	London
Model Name	Arima Weighted Ensemble	Exponential Smoothing (ETS) Ensemble	Revenue to Unit Forecast	Prophet-Based Model	PulseCast	Chronos
Model Type	Statistical	Statistical	Probabilistic	Probabilistic	Probabilistic	Deep Learning
Cost / Compute	$	$	$	$$	$$	$$$
Forecast Method	Bottom-up	Bottom-up	Top-down	Bottom-up	Bottom-up	Bottom-up
Outcomes	Quick analysis, results	Quick analysis, results	Sales goals defined for the business	Seasonal planning, layering in custom data i.e. Marketing Calendar	Seasonal planning, layering in custom data i.e. Marketing Calendar	Seasonality, retail stores (in addition to Ecommerce)
Business Type	Simple business	Simple business	Fast-growing	Fast-growing	Fast-growing	Complex, Enterprise

Rome - ARIMA Weighted Ensemble

This model blends multiple ARIMA forecasts (Auto Regressive Integrated Moving Average), each designed to capture different time-series patterns like trends, seasonality, and short-term changes.

This model is optimized for:

• Substantial sales data, with regular sales activity

• Predictable patterns that repeat over time (i.e. Seasonal Trends)

For example, this model can predict moisturizer sales as they rise in winter due to dry skin and spike again in summer when SPF-infused products gain popularity. It helps identify high-demand weeks so teams can stock accordingly.

Type of Model: Statistical Compute: $ Forecast Method: Bottom-up

Athens - Exponential Smoothing (ETS) Ensemble

This model is designed for products with stable long-term trends and seasonality to create robust predictions. This approach can be effective for products with consistent historical patterns and minimal sudden changes.

This model is optimized for:

• Minimal no zero-sales weeks or months.

• Items with at least 12 months of data.

• Items with slow, steady growth or decline in sales.

• Predictable demand with few surprises.

For example, your signature exfoliating cleanser has been slowly increasing over the past year, a notable steady growth trend, and with no surprises it stays steady week to week.

Type of Model: Statistical Cost / Compute: $ Forecast Method: Bottom-up

Bombay - Revenue to Units Forecast

The Revenue to Units Forecast Model is designed to convert projected revenue forecasts into unit sales forecasts, enabling more accurate planning. This model is particularly useful for businesses where forecasts are often made in revenue terms, but operational planning requires forecasts in physical units.

This model is optimized for:

• Businesses that forecast in revenue but need unit-level planning

• Inventory and supply chain teams requiring unit sales estimates

• Product lines with stable or predictable price-per-unit relationships

• Scenarios where price data is available to translate revenue into units

For example, if a skincare company expects to make $50,000 in sales and each skincare kit sells for $50, the model predicts they will need to prepare for 10,000 kits. This helps the company know how many toners, cleansers, and moisturizers to buy in advance.

Type of Model: Probabilistic Cost / Compute: $ Forecast Method: Top-Down

Venice - Prophet-Based Model

This model captures yearly, weekly, and holiday-based trends. It works especially well for retail products that see spikes in events such as Black Friday, back-to-school, or holiday sales. It helps you prepare for busy seasons by learning from past trends, even if some days of data are missing or sales were irregular.

This model is optimized for:

• Products with clear weekly and yearly patterns (e.g. Mondays, Black Friday)

• Holiday or event-driven sales spikes

• Sales with occasional gaps or missing data

For example, your essential t-shirt sales have a huge spike in November and December during the holiday season including a small bump in early September for back to school. The prophet model will track these patterns and recognize the holiday trends of those t-shirts from previous years.

Type of Model: Probabilistic Cost / Compute: $$ Forecast Method: Bottom-up

Hong Kong - PulseCast (New Model)

PulseCast is Moselle’s basic ensemble model. It blends three forecasting models into a single forecast using equal weighting. It evaluates patterns over time, seasonality, and messy data for a more balanced prediction.

This model is optimized for:

• Items with occasional zero-sales periods.

• Steady sales from week-to-week or month-to-month.

• Items with more than 2 years of sales history

• Capturing seasonal patterns (E.g. Holidays, End of season sales, etc).

For example, A mid‑tier skincare cream that sells steadily about 1,000 units per month with a noticeable uplift every December. PulseCast averages its three component forecasts to provide a smooth projection that preserves the holiday bump while filtering out random week‑to‑week noise.

Type of Model: Probabilistic Cost / Compute: $$ Forecast Method: Bottom-up

London - Chronos

Chronos is Moselle’s lightweight deep learning forecasting model, designed for businesses needing accurate, short‑ to medium‑term predictions with limited sales history. While it has a higher compute cost than statistical models, Chronos delivers fast, reliable forecasts that adapt quickly to stable and seasonal demand patterns.

This model is optimized for:

• Items with short history (approximately 6 - 9 months)

• Sales patterns that are simple and stable

• Fast results needed for quick decisions

• Online sales with a lot of historical data

For example, you sell a jar of all-seasoned spices in your online shop every day which gives steady and predictable sales every day. Usually, you sell the same number each day with a small increase seasonally during summer, and sales do not jump around as much.

Type: Deep Learning Cost / Compute: $$$ Forecast Method: Bottom-up