Graduation Assignment – Back-end Development

What are you going to do?

How can we reduce state queries and and increase the efficiency of our systems?

Our software maintains a digital view of retailers' stock inventory. This means that millions of items are continuously scanned, processed, and updated in our systems. Think of processes like receiving, counting, moving, or selling items. These mutations result in billions of database operations annually.

Many software applications continuously query the exact state of items to make decisions. This puts a heavy load on our databases. Consider the following scenarios:

• Assigning newly arrived stock to the correct store without counting existing stock again.
• Ensuring that already existing stock is ignored during a new shipment scan.
• Verifying whether an item is sold before leaving the store to prevent unnecessary alarms.
• Preventing e-commerce sales from being mistakenly counted as sellable store stock.

To reduce the load on our systems, we want to gain insight into how often an item's state is queried and explore ways to reduce these queries. This requires smart analysis of large datasets, potentially using data sampling techniques. Additionally, you can investigate how filtering and state checks can be performed more efficiently without always querying a central database, for example, using Bloom filters.

With your insights, Nedap can handle large amounts of data more efficiently, making our software faster and more scalable. This not only improves performance but also reduces costs and enhances the experience for our customers.