Context and challenges
This client is one of the largest companies in the world in the field of energy. It cooperates with many companies in the world to buy and sell manufacturing equipment. The detail of the exchange is stored on the raw data (SPDP: Spare Parts Data Package) and it has to be processed and cataloged. Usually this task is done manually which means it takes time and resources.
This client needed a solution that process SPDP files rapidly and efficiently. The goal of the project is to automate and improve the accuracy source of big data manipulated by many users including as well as avoiding human fault.
In this context, this client appointed Assystem to provide a digital solution as well as service support.
Assystem provided a data science approach to build a data model and analyse operational data using a specific use case, including Artificial Intelligence.
- Development of a web application enabling:
- Data capturing: OCR and read SPDP files content to extract all needed information for specific equipment (tag number, serial number, contract number, drawing...)
- Clustering purchase order contracts in different format
- Classifying the types of documents using deep learning techniques and verification throughout the testing phase
- Data cleansing or consistency: we need the output format to extract only the relevant information and compare it from different sources of input data
- User-friendly interface enabling to access and understand data easily, and the tool facilitates those actions, by automatically processing data
- Installation of the solution, the tool will be run on Assystem’s server to provide needed support to the client.
- Training of teams: preparation to run a demo with guidance documents for client.
This projets is a mix of lump sum, software delivery with some license fees and solution maintenance service to make sure that the platform is evaluative and sustainable.
- Cost savings solution and reduction of information losses thanks to improvement of speed and accuracy of SPDP process
- Skills improvements and introduction of more value in human work (from reading tasks to analyzing data and extract value)
- Improvement of relationships with the supply chain and better collaborative approach with vendors thanks to improving reporting, communication and monitoring of supplier’s activities