Problem
Customer orders arrive as several documents with detailed requirements and cross references. Changes are not always clearly marked, which makes manual checking time consuming and error-prone.
Project overview
Seal Engineering receives complex customer order documentation where requirements, references, and internal documentation must be checked carefully. The thesis solution was designed as a deterministic MVP that supports the order office by making document import, validation, mapping, and compliance checking more structured and less dependent on manual comparison.
Goal
Build a web application that can import PDF and Excel documentation, extract key information, identify relevant document types, and support compliance assessment between customer and internal requirements.
Result
The delivered MVP demonstrates a workflow from file import and validation through document mapping and compliance check presentation. The prototype is tailored to a selected customer document set.
Solution demonstration
Short video demonstration of the solution workflow, showing how the user can import a document set, validate files, extract and map values, and receive structured compliance feedback.
Solution workflow
The implementation follows the user journey from uploading documentation to receiving structured feedback about extracted data and possible mismatches.
Import files
The user uploads a document set containing order and requirement documentation, including PDF and Excel files.
Validate document set
The solution checks whether the expected file types are present and gives the user immediate feedback before the compliance flow continues.
Extract and route data
Text and values are extracted from the uploaded files and routed to document-specific mappers for Purchase Order, Material Documentation Package, MDRL, and Administrative Requirements data.
Map to domain objects
The raw content is transformed into structured objects so the comparison logic can operate on values instead of unstructured document text.
Persistent storage of structured data
The structured objects are stored in a database for future reference, versioning, and analysis.
Run compliance check
The system compares central values such as order numbers, material numbers, dates, revisions, and document references.
Present results
The result is returned to the user interface as understandable tables, checklists, status markers, and detailed document values.
Technical implementation
The source archive is structured as a .NET solution with separate Web, Application, Domain, Infrastructure, and test projects. The implementation combines a Blazor-based user interface with API endpoints, MediatR-driven application flows, document extraction adapters, database persistence, and Docker-based deployment support.
Frontend and user flow
- Blazor/Razor UI components
- Dashboard for file upload and validation
- Compliance check views with status-oriented feedback
Application layer
- MediatR commands and queries
- Document type decision and routing
- Document-specific mapping services
Infrastructure
- PDF extraction with PdfPig
- Excel extraction with ExcelDataReader
- EF Core and PostgreSQL persistence
Quality focus
- Controlled error responses
- Validation before compliance checking
- Unit and integration test projects
Thesis report
Preview the full bachelor thesis report directly on this page.