How much history do you need, and how far ahead can it forecast?

Typically two to three years of daily or weekly data. Clients running demand forecasting alongside inventory management generally plan up to 18 months out.

How is this different from the forecast our planning team already produces?

This model ranks the drivers behind demand, including external ones like weather and event calendars, and updates as new data arrives. You get the projection and the reasons for it, which is what makes it actionable.

Do we need physical inventory for this to be worth it?

No. Businesses carrying stock get the most granular version. For service and experience businesses, the value sits in smoothing demand and improving capacity utilization.

Does this replace our planning process?

No. It gives your planners a modeled baseline and a ranked set of drivers to work from, rather than a spreadsheet built on last year plus a percentage. The judgement calls stay with your team.

What data do you need to build a demand forecasting model?

Historical sales and demand data, your product, market and channel hierarchy, and explanatory variables like seasonality. Customer segment data is optional and improves accuracy where you have it.

Do we need inventory management as well?

Not strictly, the two are built independently. If you carry stock they're usually delivered together, since that's the pairing that gets you an 18 month planning horizon.

How often does the forecast update?

It updates as new sales data comes in, so the projection and the driver rankings both move with recent performance rather than being rebuilt once a quarter.

Orchestration

Runs your KAI Analytics models together so price, stock, media and customers are modeled as one system rather than seven separate answers

Book call

Explore demo

Take a quick look inside KAI Assistant

Watch a product walkthrough and see how Kleene ingests your data, builds pipelines, and powers reporting – all in one place.

No hidden fees

24/7 Support

GDPR / ISO 27001

4.6

Ask AI whether Kleene.ai is right for you

What is orchestration?

What it does

Each model answers its own question well. None of them know the others exist, so the answers never account for each other.

What is orchestration?

A layer that runs your KAI Analytics models together and works out how a change in one, like price, moves the others, like demand and stock.

What it does for your business

Gives you answers to questions like "what happens to margin if we raise prices and cut TV," instead of you having to reconcile three models' outputs yourself.

What can you ask?

What actually drove sales last quarter: price, media, seasonality or customer mix?

If we raise prices 5% on this segment, what happens to demand and to the media plan?

Which products are at risk next season once you account for both the forecast and how price sensitive their buyers are?

Where should the next £100k go: stock, media or discounting?

Which customer segments respond to TV, and are we forecasting enough stock for them?

how it works

How does orchestration work?

Connect the models

Each KAI Analytics model writes its outputs back to your warehouse as tables keyed to your IDs. Segment codes, forecast volumes, elasticity curves and channel contributions all land in one place.

Model the interactions

The orchestration layer sits across those outputs and estimates how they move together. How does a price change in one segment shift demand? How does a media cut change the forecast? Which segments are most sensitive to both?

Ask one question

With KAI Assistant on top, a CMO or CFO asks what drove sales last quarter, or what happens to margin if price goes up 3% and TV comes down 20%. The answer draws from every model at once, in plain English.

Book call

See the orchestration layer in action

Take a quick look inside KAI Assistant

Watch a product walkthrough and see how Kleene ingests your data, builds pipelines, and powers reporting – all in one place.

Connectiong models

Built on the KAI Analytics suite

Here are a few examples of how the models connect using orchestration. Any combination works, and most clients run three or more together. The more models in place, the more the layer can tell you.

Media Mix Modeling + Digital Attribution

MMM measures every channel in aggregate. Attribution resolves the digital journey to touchpoints. Fused, you get one view of media performance instead of two that disagree.

Demand Forecasting + Inventory Management

Built independently, they give an 18 month planning horizon together. Orchestration adds price elasticity and segmentation on top, so the plan accounts for who is buying and at what price.

Customer Segmentation + everything

Segment codes are available as variables across every other model, so channel contribution, forecasts and price response are all stratified by customer type rather than averaged across the base.

Lifetime Value + MMM

Most media mix models only see the first purchase. Segment based LTV multipliers feed into the MMM so channels that bring in valuable customers get proper credit, which is usually where brand activity finally gets its due.

Explore demo

Model Orchestration FAQ's

Do our teams need to do anything to keep it running?

No. Our consultants scope the interactions with you, review the first results before anything goes live and maintain the models on an ongoing basis. Your team works with the outputs.

How often does it update?

As often as the models feeding it. Most run on a weekly or monthly schedule, and orchestration recalculates whenever any of its inputs change.

Does it replace the individual models?

No. Each model still runs on its own and writes its own results to your warehouse. Orchestration sits on top and reads from them. If you only want the forecast, you still get the forecast.

How is this different from a dashboard with several charts on it?

A dashboard shows the outputs side by side and leaves you to work out how they relate. Orchestration models how a change in one moves the others and gives you one answer that accounts for it.

Is orchestration a separate product?

No. It's the layer that runs across the KAI Analytics models you already have. There's nothing extra to buy when you run two or more models with us.

Do we need every model for orchestration to work?

No. It starts adding value with two connected models and builds from there. Demand Forecasting with Inventory Management is the most common starting pair, since that's what gets you the 18 month planning horizon.

start your journey