Success Story – CYNEFY
Text extraction and detection at a globally leading electronics manufacturer
Proof of Concept for text extraction and detection in the electronics industry
Keywords
Text Extraction, Text Detection, Machine learning
Challenge
- Time-consuming comparison of datasheets during the second source process for electronic components
- Labor-intensive and error-prone copy & paste work of parameters from PDF to Excel for comparison purposes
- Utilization of internal, highly skilled resources for a monotonous and low-value activity
- Inconsistencies in decision-making due to erroneous parameters and unnecessary duplication of the same comparison process
Competencies and services
Agile project management and workshop facilitation with the customer's knowledge holders.
- Utilization of various machine learning techniques for identifying, extracting, and providing parameters from PDF files
Solution
- Analysis of existing Excel-based datasheet comparisons
- Identification of key parameters for each component class through workshops with specialists from the customer's side
- Utilization of machine text extraction and detection to extract and compare data from multiple datasheets
- Use of natural language processing techniques to clean and preprocess the text data, ensuring accuracy and consistency of the extracted information
- Comparison of extracted data fields and automated highlighting of differences
- Provision of a user-friendly frontend for analysis and decision-making by qualified personnel
- Accompaniment of the entire process through agile project management (conducted in sprints)
