Kuldeep Jiwani

Specialises In
"topological-space-creation" big-data-clustering data-science-at-scale sessionisation stochastic-periods time-sequenced-events unsupervised-ml

Researcher, Data Scientist, Data Science Architect, Performance specialist, Entrepreneur

Currently working as a Data Science Researcher and Director in Thales (Guavus) handling various Data Science projects. Along with actively exploring new techniques of Machine Learning via various research projects. Thus ensuring that Data Science could be effectively applied to real world scenarios and able to solve important problems.

The research areas are focused primarily on Unsupervised Machine Learning techniques, where we have to discover information from given data, with no available labels. The world of Supervised ML have made great advances with the advent of Deep Learning. But in majority of industries and real world scenarios, labels are hard to obtain so Unsupervised ML is the only option to proceed.

Data Science Research:

  • Researching on new Machine Learning approaches in the domain of Cyber Security, Fraud Detection, Network Operations
  • Data Geometry: Constructing topological spaces via semantic similarity between data attributes
  • Session discovery by modelling stochastic periods in time series data
  • BOD (Behavioural Outlier Detection): Behavioural models to detect Frauds via Unsupervised modelling techniques
  • NSBA (Network Service Behaviour Analytics): An unsupervised ML technique to model network service behaviours. This creates statistical Behavioural models to catch abnormal flows in the network.

Data Science Architectures:

  • Distributed AI/ML: ML over the edge and over distributed processing
  • Building large scale Machine Learning solutions over TBs of data
  • Constructing topological spaces over Big Data
  • Designing and architecting billion nodes graphs and Ontologies

Was key architect in building a JVM performance monitoring tool AD4J in Auptyma that was acquired by Oracle in 2007. Presented this tool in GIDS (Great India Developer Summit) 2008. Performance specialist in Big Data technologies - Apache Hadoop, Apache Spark, Yarn, Redis, HBase Applying Machine Learning and Machine Intelligence to Big Data problems to do analytics at large scale