Case studies

AI implementation case studies: examples of process automation, machine learning and systems integration

Each case study shows the industry, the problem, the solution and the result. We do not publish client names.

01 - Area

AI process automation

Language models, AI assistants and computer vision take over work with documents, correspondence and images.

cm-opti projectInsurance · NLP · Document automation · Regulated industry

Claims correspondence classification at an insurance company

An NLP model classifies an insurer’s claims correspondence in a GDPR-compliant environment. 25% fewer documents in the main queue, 90% accuracy, 30% smaller backlog.

cm-opti projectConstruction · Project management · RAG · Regulations and compliance

AI assistant for German construction regulations (HOAI and VOB)

An AI assistant answers questions about the German HOAI and VOB regulations and cites the legal basis. Getting to an answer takes 10 minutes instead of 1.5 hours.

Team experienceRetail · Computer vision · Stock control · Automation

Computer vision: matching shelf stock to system inventory

Computer vision compares shelf photos with system stock and flags discrepancies before they turn into stockouts or wrong orders.

02 - Area

Machine learning and forecasting

Machine learning models and data analysis support decisions on inventory, sales and product development.

Team experienceDistribution and logistics · Machine learning · Demand forecasting · Optimization

Demand forecasting and distribution optimization

ML models forecast demand at the item level, and optimization plans distribution while accounting for logistics costs, lead times and warehouse capacity.

Team experienceData and BI · Reporting · Power Query · KPI

Reporting automation: from 2.5 days to 2 hours

Standardizing and automating management reporting. Report preparation cut from 2.5 days to 2 hours with Power Query and clear KPIs.

Team experienceRetail · Recommendation system · Market basket analysis · Collaborative filtering · Cross-selling

A recommendation system built on hundreds of thousands of receipts

Analyzing hundreds of thousands of receipts shows which products are bought together. The patterns feed cross-selling, inventory planning and online product recommendations.

Team experienceProduct development · NLP · Clustering · Voice of customer

Forum content analysis: a new customer segment and product development

Clustering and analysis of posts from special-interest forums. A map of user expectations for a new customer segment and the development of a higher-margin product.

03 - Area

Systems integration

Systems exchange data on their own, and staff only approve the result.

cm-opti projectLogistics and freight forwarding · Systems integration · OCR · Document automation

Systems integration and automatic document capture in logistics

Capturing data from scans and emails, automatic entry into the TMS and data exchange between systems. Manual retyping is kept to a minimum.

Confidentiality and ownership

We do not publish client names or project details. We run every project in a secure environment, on the client side or in the cloud, in line with legal requirements and the GDPR. The rights to the code remain with the client, and the client’s data stays in its environment.

Not every process needs automation, and some companies need to get organized before they need technology. If you want to know where your company stands, describe your process. We will get back to you by the next business day and tell you plainly whether we think it makes sense.