NXTGEN BIKE
AI & Data
Semester programme:Minor Data Driven Business Lab
Project group members:Nikola Milkov
Anastasia Grabina
Dayeon Lee
Project description
The project focuses on answering the following research question:
How can a validated proof of concept for a data pipeline that extracts the right data from the MKG ERP system via an API, expands it with an LLM, and sends it to a simulation model developed by TU/e be created?
Context
This work is conducted as part of the NXTGEN Smart Industry 06 — Smart Supply Networks project, which aims to study how interaction processes within supply chains can be modelled and improved using operational data. The project is centred around the Smart Connected Supplier Network (SCSN) concept, designed to create an information-sharing environment to enhance process efficiency.
The supply chain itself is viewed as a network of all parties involved, directly or indirectly, in fulfilling a customer request: from manufacturers and suppliers to carriers, warehouses, and end consumers. The effectiveness of this network is determined by the interaction of logistical and cross-functional factors, such as production, inventory, and transportation. Within this project, information exchange serves as the key tool for synchronising these elements.
Results
The results of the project are:
- Validated applied research on the topics: Supply Chain Management, Enterprise Resource Planning, Application Programming Interface, Synthetic Data, Exploratory Data Analysis, Prompt Engineering, and Simulation Models.
- Deep understanding of the MKG ERP system and its data structure.
- Development of a reactive Marimo Notebook that serves as the validated proof of concept for a data pipeline. It visualises the data extracted from the MKG ERP system, traces orders in the supply chain, and uses a built-in synthetic data generator to expand the data.
- Data quality analyses and visualisations.
- Prompt engineering for creating large synthetic datasets.