Industry
Technology / Semiconductor Industry
Work Area
Data Management, System Integration & AI-Driven Process Optimization
Project Scope
Scalable, global solution for over 300 users, with over 10 years of continuous development
Initial Situation
Employees working around the world had to access data via a decentralized SAP solution using a VPN in order to prepare quotes. This severely limited performance: each click and query resulted in wait times of several seconds. Furthermore, the standard version of SAP was not designed to display all information relevant to a customer call in a high-performance, consolidated manner on a single screen. This led to inefficiencies and frustration among the sales teams.
Objective
- Full automation: AI is used to generate all product data (product recognition, title, description, color, size, material, and price) from a single product photo.
- Increased efficiency: Drastic reduction in processing time per item.
- Omni-channel publishing: Automatic, cross-platform listing of items (multi-channel) with the option for manual intervention (human-in-the-loop).
Success Factors
- Pragmatic MVP approach: The decision to forego complex full API integration in favor of a daily export ensured a quick project launch.
- High scalability: The system now effortlessly processes 4 to 5 million customer records and 20 to 30 million quotes while maintaining consistently high performance.
- Human-in-the-loop: The targeted involvement of employees in AI validation guarantees the highest data reliability.
- User-centricity: Consistent adaptation of the user interface to department-specific needs.
Description of the Case Study
1. Pragmatic SAP Integration & Data Structuring
Instead of a complex, direct full SAP integration (API), which would have posed significant technical and time-related challenges, an efficient daily export (CSV) of customer and master data was implemented. This data was not only imported but also structurally optimized for daily sales operations and migrated to a custom-built database.
2. High-performance data retrieval
The new architecture makes it possible to search through millions of data records in a fraction of a second. Even complex individual queries are guaranteed to take less than 1 to 2 seconds.
3. Group-wide transparency
In the next phase of the project, logic was integrated to automatically identify and link subsidiaries and global holding companies (at the group level). This ensures that quotes are presented in a consistent and transparent manner across all companies.
4. Data Validation & AI-Powered Deduplication
Manual, error-prone validation processes using Excel have been replaced. Today, AI models automatically detect duplicates in the system. Using a human-in-the-loop approach, employees review critical cases, while 100% unique matches are accepted fully automatically. To maintain SAP as the “single source of truth,” the cleaned data is fed back into the SAP system via a secure export-import process.
Benefits
- Increased productivity: Global use by nearly 300 employees with user interfaces customized for each department.
- Data quality: Significant improvement through system-supported validation, intelligent data consolidation during mergers, and AI-driven deduplication.
- Employee satisfaction: The high-performance solution enables smooth workflows that cannot be replicated at this speed even with newer SAP versions (such as S/4HANA).
- Data security: Indirect access (read-only with rollback export) ensures that core systems remain protected.
Opportunities
- Increased efficiency and revenue
- Enhanced employee experience
- Greater employee retention through modern tools
- Highest data quality for strategic decision-making
Risks
- Performance Risks Amid Growing Data Volumes
- Security Risks: Global Access Requires Strict Access Control
- AI Misjudgments in Data Integration
- Full SAP Integration Would Likely Have Prevented the Project from Ever Getting Started
In short:
A high-performance database solution developed specifically for quote generation overcomes the performance bottlenecks of decentralized SAP systems. Thanks to pragmatic data structuring, AI-powered data cleansing, and a strong focus on the user, approximately 300 employees worldwide are working more efficiently than ever before. The solution has been in successful use for over ten years, scales continuously, and has long since become indispensable to the workforce.
Future-proofing your company
