Machine Learning in Supply Chain: case study Plum x Flowlity

Machine Learning is increasingly used in Supply Chain planning, but many discussions remain theoretical. Companies hear about AI forecasting, automated replenishment, or inventory optimization, yet it can be difficult to understand what these technologies actually change in day-to-day operations.

The experience of Plum Living provides a concrete example. As the company grew, its Supply Chain became more complex: more products, more suppliers, and increasing demand variability. Planning processes that once worked in Excel were no longer sufficient to manage inventory efficiently.

This case study explores how Plum Living introduced Machine Learning into its planning process, moving from manual inventory management to data-driven forecasting and replenishment decisions. The transformation improved visibility across the supply chain and significantly reduced inventory levels while supporting the company’s growth.

Plum Living: a fast-growing digital-first furniture brand

Founded in Paris in 2020, Plum Living specializes in customizable kitchens, wardrobes, and bathrooms sold through an online model focused on design and personalization. The company quickly built a strong brand presence across Europe and attracted a large online audience.

At the time of the project, the company operated with:

• around 45 employees
• approximately 1,000 SKUs, combining make-to-stock and make-to-order products
• around 10 suppliers supporting production and sourcing
• two warehouses managing inventory flows

Like many fast-growing digital brands, Plum Living faced a common operational challenge: scaling its supply chain while maintaining efficiency and service levels.

As demand increased and the product catalog expanded, the limitations of manual planning became more visible.

When growth exposes the limits of manual Supply Chain Planning

Before implementing an AI-driven planning system, Plum Living relied heavily on spreadsheets to manage inventory and replenishment.

While Excel provided flexibility in the early stages of the company, it gradually became difficult to maintain accurate and reliable planning as the business expanded.

Several operational problems appeared:

High inventory levels

Inventory was unevenly distributed across product categories, tying up working capital and increasing storage costs.

Limited visibility on future demand

The planning team lacked reliable forecasts, making it difficult to anticipate replenishment needs.

Manual replenishment processes

Inventory planning relied on manual calculations and spreadsheet updates, which made decision-making slow and error-prone.

Weak supplier visibility

Suppliers had little visibility into future demand, which complicated coordination and production planning.

Stockouts not systematically measured

Shortages were not always tracked or analyzed, preventing planners from identifying recurring issues.

These challenges are typical of organizations that grow faster than their planning processes. The supply chain becomes reactive, with planners spending most of their time correcting problems rather than anticipating them.

Introducing Machine Learning (ML) into Supply Chain planning

To address these issues, Plum Living decided to modernize its planning process using an AI and Machine Learning-driven Supply Chain planning platform – Flowlity.

The goal was not simply to improve forecasting accuracy, but to support better planning decisions across inventory management and replenishment.

The system analyzes historical demand patterns, lead times, and operational data to generate planning recommendations and improve visibility across the Supply Chain.

Once implemented, the platform provided several capabilities:

• demand forecasting based on historical patterns and variability
• automated replenishment recommendations
• inventory optimization based on service level targets
• improved coordination with suppliers through shared planning horizons

This approach allowed planners to move away from manual spreadsheet management and focus on monitoring exceptions and adjusting planning strategies.

These capabilities are part of a broader shift toward Supply Chain automation, where companies progressively replace manual planning with scalable tools designed to support growth. For a deeper look at this evolution, explore our guide on the most scalable Supply Chain automation tools to support growth and retail expansion.

A fast deployment focused on operational impact

The implementation was designed to deliver value quickly. The project followed several phases:

Supply Chain analysis

The first step involved mapping Plum Living’s operational processes and identifying the main planning challenges.

Data synchronization

Operational data from Plum Living’s systems was integrated into the planning platform, allowing the models to analyze historical patterns.

Model calibration and training

Machine Learning models were trained using the company’s historical data to improve forecasting accuracy.

Operational testing and deployment

Planning recommendations were tested with real use cases before being deployed in daily operations.

The full deployment took approximately three months, covering two warehouses and around 630 SKUs.

This relatively short implementation timeline allowed the company to quickly move from manual planning to a more automated, data-driven approach.

From forecasting improvements to better inventory decisions

One of the key changes introduced by the new planning system was the ability to generate a longer-term view of demand and replenishment needs.

The platform produced a nine-month replenishment horizon, which could be shared with suppliers to improve coordination.

This longer planning horizon helped teams anticipate supply constraints earlier and reduce uncertainty across the upstream supply chain.

At the same time, automated replenishment recommendations simplified the daily work of planners. Instead of manually recalculating inventory needs, planners could review suggested actions and focus on exceptions.

The system also helped improve visibility across product categories, making it easier to rebalance inventory levels and adjust stock policies. This places Flowlity as one of the best Supply Chain visibility tools for the retail industry.

Results: lower inventory and smoother planning processes

The impact of the transformation became visible quickly.

After deployment, Plum Living achieved a 21% reduction in inventory levels, freeing up working capital while maintaining service levels.

Additional improvements included:

• automated and streamlined replenishment processes
• improved supplier visibility through shared forecasts
• smoother integration of planning into daily operations
• significant time savings for planning teams
• better support for new product launches thanks to improved forecasting

Long term results also show an overall inventory value reduction of around 38% as planning strategies continued to improve.

For operational teams, the transformation had a direct impact on daily work: “When Flowlity arrived, it truly changed my day-to-day life” explains Axel Moulhiac, Logistics Manager.

What this case study shows about Machine Learning in Supply Chain

The Plum Living experience highlights several important lessons for companies considering ML in Supply Chain Planning.

First, Machine Learning creates the most value when it is connected directly to operational decisions. Improving forecasting alone is not enough; the real benefits appear when forecasts drive better inventory and replenishment strategies.

Second, implementation speed matters. Many companies hesitate to adopt new planning technologies because they expect long and complex deployments. In this case, the transition from manual planning to AI-supported planning was achieved in only a few months.

Finally, technology adoption must support planners rather than replace them. Machine Learning helps process large volumes of data and detect patterns, but human expertise remains essential to interpret results and adjust planning strategies.

‍

Similar transformations can be observed in other organizations undergoing Supply Chain modernization. For instance, in this case study on Camif’s digital transformation, you can see how another retail player leveraged digital tools to improve planning and operational performance.

‍

Download the Plum Living case study

This article provides a detailed overview of Plum Living’s supply chain transformation.

For a shorter visual summary of the project, including key figures and operational highlights, you can download the complete case study filling out the form below.

‍