Optimizing a Swiss SME's inventory management with Business Intelligence (BI)
Business IntelligenceChatbotAI in business

Optimizing a Swiss SME's inventory management with Business Intelligence (BI)

Discover how a Swiss SME reduced overstocking by 15%, virtually eliminated stockouts, and saved 6 hours per month by automating data collection, daily reporting, and stock alerts through a simple BI solution and conversational chatbot.

3 min readPublished on June 12, 2025

A Swiss SME manages daily inventory of raw materials to supply its production. Previously, managers ordered large quantities to benefit from economies of scale. In practice, these large volumes resulted in unused surplus in the warehouse and did not actually provide financial advantages. Moreover, planning relied on a simple Excel spreadsheet updated weekly, without precise data on daily consumption. This approach led to both excess stock and risks of stockouts at critical moments for production.

Challenges

Limiting overstock

Reduce immobilized volumes to free up cash flow and storage space.

Preventing stockouts

Avoid any production interruption and delivery delays to customers.

Obtaining an up-to-date view

Have reliable information available at all times on stock levels and consumption trends.

Project objectives

Over a two-month pilot period, deploy a simple BI solution allowing to:

  1. Automate data collection
    Centralize every evening:
  • Raw material receipts and shipments
  • Daily consumption by the workshop – Group this information in a single file, without manual intervention.
  1. Generate a daily report
    Send every morning a document (PDF or Excel) indicating for each material:
  • Remaining stock (in units or kilos)
  • Average consumption of the last seven days
  • Number of days remaining before reaching the critical threshold
  1. Issue automatic alerts
    When estimated stock drops below 20% of normal level, an email is sent to the purchasing manager:
    Alert: Material B stock below 200 units. Please plan an order today.
    This notification avoids any surprise on the morning when production might be interrupted due to lack of material.

  2. Offer simple access to data
    Connect a chatbot to the BI solution to allow teams to ask questions in natural language (for example: "Which materials should be monitored this week?") without having to go through reports manually.

Deployment

Pilot phase (2 months)

Selection of critical materials

Identify three high-impact references (the most consumed).

Implementation of collection script

A discreet tool groups consumption and receipt data every evening.

Calculation of average consumption

Estimate, in a simple way, the average daily consumption over a week.

Automatic sending of reports

Daily sharing with purchasing and production managers.

Adjustment of alert threshold

Weekly meetings to refine sensitivity (expanding threshold to 25% to reduce false positives).

Transition to operational mode

  • At the end of the pilot phase, the process was generalized to all managed materials.
  • Every morning, managers now consult the report before planning orders or revising planning. They also have the possibility to directly chat with the database thanks to the integrated chatbot.

Results obtained

Reduction of overstock

In three months, the total volume of stored materials decreased by 15%, freeing up about 25,000 CHF in cash flow and warehouse space.

Disappearance of unexpected stockouts

Before the project, a stockout occurred on average twice per quarter. After deployment, only one minor incident occurred in four months, resolved in a few hours thanks to the morning alert.

Time savings on data preparation

Manual updating of the Excel spreadsheet required 8 hours per month. With the automated report, this time dropped to 2 hours (checks and adjustments), representing a net gain of 6 hours per month.

Better visibility for the purchasing team

The team spotted a recurring consumption peak for a material at the beginning of the month. By anticipating and placing the order two days earlier, they avoided any activity interruption.

Conclusion

This example shows that a BI project, even simple, can quickly generate value:

  • By automating data collection, the Swiss SME gained reactivity and controls its stock levels.
  • Morning alerts helped reduce costs related to excess volumes and avoid stockouts before they disrupt activity.
  • Success with three materials convinced the team to extend this approach to the entire inventory management.
  • Adding a conversational chatbot reinforced data accessibility: teams can interact with key information in natural language, without having to manipulate complex tools.

Long term, a global dashboard will allow tracking, in addition to materials, key indicators such as profitability per product and delivery times, while maintaining a clear interface without technical jargon.

Do you want to optimize your inventory management with AI? Our AI consulting and development team is at your disposal to design and deploy a BI solution adapted to your needs.

Written by

Hugo Desbiolles

Hugo Desbiolles

AI Consultant

Get in touch & let's make something!

Want to go further?

Let's discuss your challenges and define what AI can do for you.