Synthetic data generation
At Needl, our mission is to organize and stitch your information to make it universally accessible and useful. Knowledge workers today are inundated with massive amounts of data via multiple communication apps and devices resulting in huge efforts to save, organise, retrieve, and make sense of data leading to productivity loss. Needl aims to unbundle your data across apps & devices into a single repository for both structure and unstructured data across private and public sources. A seamless experience of all your data in one place, securely backed up with a host of cloud computing processes on tap and user defined interfaces built to analyse and share – transforming the way you and your team work and collaborate!
As our engineers continue to develop new features (especially ML-related), we needed a way to test those features against user data. But since the privacy of every user’s data is non-negotiable, we cannot directly use Production data. We wanted a way to generate synthetic data from a snapshot of the Production data and then test our features reliably against this synthetic data.
Slides outline (UPDATED 13 Apr) https://docs.google.com/presentation/d/13ObjJyCl2a38yUud-pxVx34plo5t50zZb7l0K3sNB8Y/edit?usp=sharing