A collaborative effort between the University of Pretoria and SITA, to develop a java platform that would support the implementation and maintenance of E-Government services. The tool strongly supports re-usability, and in so doing exploit variability modeling to optimize the creation of these services. The project is currently an open source initiative and can be found on github,  https://github.com/spillay/eTool

Variability Support in Domain-Specific Language Development
Edoardo Vacchi, Walter Cazzola, Suresh Pillay, Benoît Combemale


BigData Analytics

Currently collaborating with Jaasuz, which is a start-up company that provides advanced data mining algorithms and software technologies for text anaytics to detect/predict various types of behavioral patterns for enterprise risk management, ecomm surveillance, social media (e.g, Facebook, Twitter, craigslist) and
Internet (e.g., email, blogs).

For example, such solutions can detect many forms deceptive behavior from data include: lying, impersonation, gender falsification, malicious sexual intent, falsifying profile information, and coded communications.

6th International Conference, SLE 2013, Indianapolis, IN, USA, October 26-28, 2013. Proceedings

Suresh Pillay


Publications and collaborations

TEChnology architect

Java platform

Domain Specific Languages (DSLs) are widely adopted to capitalize on business domain experiences. Consequently, DSL development is becoming a recurring activity. Unfortunately, even though it has its benefits, language development is a complex and time-consuming task. Languages are commonly realized from scratch, even when they share some concepts and even though they could share bits of tool support. This cost can be reduced by employing modern modular programming techniques that foster code reuse. However, selecting and composing these modules is often only within the reach of a skilled DSL developer. In this paper we propose to combine modular language development and variability management, with the objective of capitalizing on existing assets. This approach explicitly models the dependencies between language components, thereby allowing a domain expert to configure a desired DSL, and automatically derive its implementation. The approach is tool supported, using Neverlang to implement language components, and the Common Variability Language (CVL) for managing the variability and automating the configuration. We will further illustrate our approach with the help of a case study, where we will implement a family of DSLs to describe state machines.

In collaboration with Prof R. Chandramouli from Stevens Institute of Technology and Suresh Pillay from WorkSpace Africa (Pty) Ltd, has partnered with Spectronn a disruptive start-up company based in New Jersey (USA) to provide intelligent spectrum harvesting and management. This allows us to provide reliable and secure mobile broadband connections over multiple bands within one wireless network or simultaneously across multiple heterogeneous wireless networks.  Reducing cost for wireless carriers, supporting urban and rural mobile broadband, robust public safety communications in emergencies, secure military communications and more.