FAQ - Old page

Flowlity's approach is step by step, we first go to one company then we go to suppliers and customers. With no business partners using Flowlity, the better use of in-house available data already allows for dramatic inventory reduction thanks to:

• Higher accuracy for demand forecast on finished goods

• Higher accuracy on raw materials/components consumption forecast

• Evolutionary placement of inventory: Flowlity recommends higher levels of inventory in points in time where uncertainty is high and lower levels where uncertainty is low.

• Frequent update of inventory levels recommendations: where usually business review their inventory policies annually, Flowlity allows to updating them on a regular basis taking into account recent events

Flowlity is the first planning solution that is improving as more and more customers are using it.

DDMRP is a methodology that greatly inspired Flowlity. Some of the principles are similar : Consumption sensing instead of MRP calculation, Visual indications to identify priorities, Min-Max replenishment strategy. Flowlity is a software not a methodology. The main difference resides in the continuous automation and improvements of recommendations and the fact that Flowlity as an E2E approach to the planning when DDMRP has still a silo-ed view of planning.

We run planning clinics to measure accurate ROIs. We answer the question : what would have happened if Flowlity was used for the past year?

We use actual data and simulate replenishments decisions.

In average, we have seen inventory decrease in between 20 to 60% depending on the company.

Yes, this is actually what we do. We do forecasting for the sake of placing inventories and replenishments. It only makes sense at the lowest level.

We can also do forecasting at higher levels in the DEMAND module.

Our main competitor is the status quo. We are actually doing something completely new that has never been tried before. We need to explain and sell the concept first.

Main other software vendors are selling inventory optimization solutions. But Flowlity's concept of intelligent 3rd party yields to significantly better results.

Yes! It works well with consignment stocks. Saint-Gobain and Plansee are prefect consignment stocks usecases.

The problem we solve is quite robust to data quality. Bad data means higher uncertainty which means higher safety stock.

Better data will in return mean less need of a high safety stock. During planning clinic and implementation project we provide recommendations to improve data quality.

For lead time, we predict the lead time in between ordering and reception of the goods. This is the replenishment lead time.

Nope. Flowlity pays for itself almost immediately. Subscription costs • ~.042% of average past inventory value paid monthly • You can cancel your subscription at anytime

ROI in less than 3 months. We can compute it precisely by doing a Planning Clinic.

If the scope is big, we can consider it on a case by case basis.

From 2k to 20k€ depending on the scope, the complexity. It is never free as we want a clear commitment from the prospect. We provide a highly qualitative ROI analysis.

You should go for a Planning Clinic. We will give you the answer to the following question : How much would you have saved if you had used Flowlity for the past year.

In general 6 weeks, 3 for IT integration and 3 for testing and parallel run. It can be longer for big and complex projects.

To onboard you business partners, we offer a lightweight access to the application. For example, if you are using the PLANNING module, we can propose to your suppliers access to the DEMAND module that will provide to them a way better sales forecast.

The solution is based on latest web technologies (Javascript, Python, ...) and the machine learning algorithms on Python libraries (Tensorflow, SciKit learn, ...)

Azure datacenter in France.

Our biggest customer so far as 100k+ SKUs. We have built the solution to scale on big volumes.

Yes, we have semi-interfaces with standard SAP tables provided in the IT integration guide.

We use a divers set of machine learning algorithms : Trees, Deep learning, Neural networks, ... We combine them together in a technique called ensemble learning to achieve the highest possible forecasting accuracy.