Automate and streamline data extraction from customs docs
Register for free Trial
Problem to be Solved
Streamline and automate the extraction of information from transit declaration documents. Enhance efficiency, ensure data accuracy, and promote compliance with regulations. We aim to support SGS in delivering faster, error-free services to customers while also supporting their sustainability goals.
Problem Statement
- How to handle a large volume of documents with various formats? The documents have many fields and don't follow a specific order, and some documents are hundreds of pages long.
- How to reduce the review effort? Only review forms that are suspected to be inaccurate.
- How to handle multiple languages? (European countries)
How QAI solution solves the problem?
- The system consists of a mail bot, a portal, and an AI engine.
- The Mail Bot can receive orders from users. Once it receives an order, it sends the information to the portal for processing.
- After the order is processed (via OCR), the mail bot sends a notification to the user about the processing results.
- In cases of data errors or when the confidence level is low, the mail bot sends an email to the operator for review.
- Portal allows files to be uploaded, extracts information, and exports data to SAP.
- The AI system, integrating OCR, NLP/LLM, and various image and text processing models, enables recognition of many types of invoices in different languages. The system can process new formats/layouts without requiring model retraining.
- The AI system also assesses the reliability of extracted information, highlighting invoices that need operator review.
Why the problem was not solved before?
- Too many forms.
- Complex forms.
- Diverse languages.
- Too much review effort.
Results from the solution
- Average accuracy: >90% for common forms, over 75% for new forms.
- Applicable to 15 countries, reducing manual effort (75-80% reduction in processing time)
3/4/2025
NEW
Apply AI-Model to optimize the identification and classification of products (based on analysis, product details, unit price, unit of measurement, etc.)
NEW
"Use a large language model (LLM) to identify descriptive features from a reference image.
Based on these features, users can customize specific attributes and use the customized information to search for or generate images with similar characteristics."
NEW
Crawl data from websites and use a large language model (LLM) to extract and summarize information that aligns with the user's needs.
QaiDora Products
QaiDora draws inspiration from the myth of Pandora’s box—a symbol of unexpected possibilities and hope. For us, AI models are like modern Pandora’s boxes, holding untapped potential to turn challenges into opportunities. At QAI, QaiDora serves as an ecosystem of AI products designed to drive innovation and deliver competitive advantages.
Trusted by
