SynthDeep

SynergyTop Successfully Created UI Design For A Synthetic Data Generation Tool

Our world is run by data today. And synthetic data is especially important:

• As a stand-in for test datasets of production or operational data
• To validate mathematical models
• To train machine learning models

To help one of our esteemed clients generate synthetic data using some raw data, our team was chosen to build the web UI for the synthetic data generation tool.

Client’s Goals

The key goal of this project undertaking was to provide a web UI to users which allows them to generate synthetic data from raw data.

The Challenge

The key challenge involved in creating such a simple, streamlined, yet powerful solution was the creation of a user-friendly web UI that makes the solution reliable and easy to use for the end-users.

Tech stack

The following technologies were leveraged for creating the UI for the synthetic data generation solution:

• React-based templates
• S3 bucket for data storage
• ML Pipeline

Here’s what the architecture of the project looks like:

The solution

The solution lets users enter raw data and quickly download synthetic data for various purposes.

The solution was designed based on the following user flow:

In case there is an error in generating synthetic data, an error log is generated and displayed.

The results

Based on the successful completion of the 1st phase of the solution (i.e. Web UI creation), the client entrusted SynergyTop with the task of implementing RBAC (Rule-based access control) to the solution in the next phase of the project.

In this second phase, our team added Admin roles and write and read roles as opposed to all users being allowed access to all functionalities in the 1st phase.

With that, our team completely handed over the solution to the client who is now leveraging the solution for various synthetic data generation needs.