Screenshot image source Unsplash.com. Yara logo is a trademark belonging to Yara International ASA.
About client
Yara fosters understanding to nutritionally sustain the planet while protecting the ecosystem. Supporting the vision of a world free of hunger and a planet that's respected, Yara advocates for a strategy of sustainable growth, promoting climate-aware farming nutrients and zero-emission energy solutions. Yara's purpose is geared towards developing a nature-friendly food outlook that adds value to our customers, investors and the wider society, and creates a greener food value chain.
Business Challenge
Yara is facing challenges in managing and streamlining its generated data. This results in slower decision-making, inefficient processes and fragmented information about order status. As any corporation, Yara utilises many different systems used by different business domains. This brings complexity of integrating data generated by those systems and utilising in operational layer:
Diverse Systems: Many different systems for various operations.
Lack of Sales Visibility: Inability to track sales progress effectively.
Order Process Uncertainty: No clear insight into the order processing stages.
Isolated Systems: Lack of communication and integration among existing systems.
Communication Challenges: Relying heavily on email and phone for communication. Information kept in emails and Excel spreadsheets.
Limited Information Sharing: Important information primarily resides in employees' heads, causing knowledge silos.
Technical challenges
Primarily there three major challenges:
Single source of truth: Identify different data source systems and define what datasets are relevant to cover our use cases and strategy.
Reusability of datasets: Generalise the necessary datasets so it could be self explanatory reusable beyond our known use cases.
Make solution business people friendly: Select and setup tools that would allow early solution adopters and champions within the organization tweak and adjust processes however they see fit.
Solution
We went through multiple stages of defining use cases and operational layer tied to respective use cases. Then applied Data Mesh principles, identified data sources and transformation methods. Built POC and tested it against use cases.
Understanding Operations: Gain a comprehensive understanding of the company's operations.
Applying Data Mesh principles: Implement data mesh principles to optimize data management.
Identifying Data Sources: Identify and catalog all necessary data sources within the organization.
Data Transformation: Transform and prepare the data for use in data products.
Proof of Concept (POC) - COSY System: Develop a proof of concept (POC) for the COSY system using Power Apps.
Testing Management Process: Conduct POC testing to validate the desired management process by manually inputting information.
Results
Data Mapping Completed: Successfully completed data mapping process.
Data Product Segmentation: Data has been categorized into distinct data products.
Structured Data: Data is now organized according to the required Data Verse configuration.
Data Readiness: Data is prepared for independent use within the company.
Efficient Manual Work: Methodical manual work and data entry processes established.
Document Template Creation: The client has initiated the creation of document templates.
Sales Visibility: Now have visibility into the ongoing sales status.
Information Centralization: Information is now centralized within dedicated platforms for better accessibility and management.
Future capabilities
Automated Data Flow: Implement automated data movement processes for greater efficiency.
Enhanced Interconnectivity: Foster seamless integration between business applications to ensure integrity and coherence.
Advanced Data Visualization: Develop capabilities for advanced and insightful visualization of analytical data.
Decentralized Data Responsibility: Empower departments to take ownership of their data products and management.
Data Science Integration: Incorporate data science techniques for deeper insights and predictions.
AI/ML Integration: Explore the integration of Artificial Intelligence and Machine Learning for data-driven decision-making.
