Case Study


KnowYourDay is an Employee Performance Monitoring System which logs the activities of a user while predicting user behaviour and activity classification with Neural Network based Artificial Intelligence supported by algorithmic analysis. The platform creates detailed reports based on empirical observations and metric analysis.

For more Information visit

Critically, the system uses aggregate data only to avoid the collection of any sensitive information.

Analysis can be grouped by projects, employees, departments, teams and tags. The goal of the platform is to support managers to understand the working patterns of their team, with more transparent & flexible working systems, enabling business transformation insights to increase the productivity.


The Process



To collect the system logs, DNS, keystrokes, the file activities, time and application switch activities of a user in an encrypted cloud environment, we needed a lightweight endpoint for Max and Windows platform.

The platform must also perform behaviour analysis and activity classification based on the Neural Network algorithms to create an alert system triggers alerts when targets are met and when unusual behaviour is noticed in employees.

There was a need to have a centralised dashboard system for moderating endpoints, creating advanced real time timeline charts, performance comparison charts and billable reports. Also a remote push installer by which endpoints can be seamlessly installed over a network without a manual intervention.


  • Getting root access and providing enough privileges to the endpoints for collecting DNS records.
  • Despite collecting various logs from the system in real time, CPU usage and Memory of the software must be optimized.
  • Auto updating the endpoints when an update is released without crashing or creating downtime for the app running in the background.
  • Maintaining and syncing the local database with the cloud storage whenever the internet connection is resumed.
  • Seamlessly install the end points across any scalable network through a single push button.
  • Understanding what the team is working on via Neural Network based categorisation of complex data with AI logical engine.


The whole architecture is based on a non-blocking distributed code based with socket connected data flows (ES6, NodeJS, MongoDB and REDIS & Swift for Mac). We collected the raw logs from the user machine and aggregate it to a usable format.

We had to do data reduction to optimise the data and fetch the data faster from the database. Besides, we also used REDIS to view the user’s live log and to collect the user’s log continuously without any downtime.