Privacy Preserving Analytics, from a POC to an MVP

Purpose and goal:

The project aims to build features in our product that make our solution applicable to new customers who belong to similar market segments to the company we are working with (a Swedish telecommunication company). The tasks to be developed in this stage are conceived to facilitate the transition of our proof-of-concept(POC) with the Telco giant into a minimum viable product (MVP).

We have successfully addressed all the project tasks, which positions DPella in a better position to acquire more customers. This project´s results will also help demo our technology to prospective customers. Expected results and effects:

The results of the different tasks have advanced the technology of the company in many directions. It has increased and improved our code base. We can now quickly program the generation of datasets (task 1) and customize analytics (task 3) while being sure that our solution is applied correctly (task 2). We have also envisioned a suitable manner to deploy our software to our customers (task 4).

All of these results make it smoother to demonstrate our technology to prospective customers and subsequent acquisition. Approach and implementation:

Task 1: Create a programming language to generate synthetic (fake) data. It is often a no-go for potential customers to send data to DPella to test our technology. Task 2: Ensure that Differential Privacy is applied correctly by design via programming language techniques. Task 3: Offer a high degree of customization on histograms (a basic analytic).

Task 4: Deploy DPella´s software on customer infrastructure with the guarantee that their data cannot be leaked.