E-Commerce Platform


Planogram compliance for detecting missing or misplaced items. A retail store frequently has to cope up with demand & much larger stores need to employ staff just to check up on inventory, products placed on the shelves & get items in place misplaced by customers, a very frequent occurrence & annoyance for storekeepers. With that taken into account, we have a solution to get instant updates on if a shelf is empty, the item is missing & our end-to-end system gets real-time updates from inventory linked to the database & assigns someone instantly to fix the situation.


  • Computer Vision
  • Deep Learning
  • Machine Learning


  • Python

Key Technical Challenges:

  • Having varying products on the shelves, we had to make sure we make generalized model
  • Data provided to us did not cover all the use cases & we had to get new data & annotate them in order.
  • Establishing a streamline pipeline with a Deep learning model was a bit difficult.

Business + Technical Points:

  • Needed real-time updates & notifications from the store
  • Role-based access to the system & Modification of positioning of shelves should not affect the detection
  • Surveillance quality was good but modernization may happen in future, the system should cope with it.
  • Resource consumption in terms of model should be optimal & should be scalable when need arises for expansion.