• Home
  • Print CV
  • Contact me

Career Objective

I work on the intersection of Artificial Intelligence and Software Systems. My research concentration is Machine Learning for software reliability. Over the years, I developed tools for program comprehension, vulnerability detection, and modeling program changes. My dissertation's primary focus is on improving the software development pipeline(i.e., building better developer assisting tools) with Machine Learning and Deep Learning. I am interested in work problems where I can apply the knowledge I gained throughout my Ph.D. career as well as learn new things to embrace new challenges.

News

  • PLBART is accepted at NAACL'21. [code] Pre-trained models coming up soon.
  • Passes the Ph.D. candidacy exam. The abstract of the candidacy can be found here
  • CODIT got accepted at IEEE Transaction of Software Engineering.

Research

I am a PhD candidate at Columbia University working with Professor Baishakhi Ray in ARiSE lab.


Current Projects

I am working to build a Vulnerability detection system in source code to detect Vulnerabilities during the Software Development coding phase automatically. We use different graphs in code (e.g., Control/Data flow/dependency graph), which play a vital role in the vulnerable characteristics of source code. Different problematic interactions of code elements in these graphs are mostly manually identified and (often) hard-coded in vulnerability identifiers. Our effort is to reduce such developer burdens by learning such interactions in a data-driven fashion.


Publications

Work Experiences

Google LLC.Sunnyvale, CA
May 2019 - August 2019

PhD Intern

Worked in designing ML based tool for analyzing security and privacy.

Fujitsu Laboratories of AmericaSunnyvale, CA
May 2017 - August 2017

Research Intern

Supervisor : Dr. Ripon K Saha
Worked with AI based program repair tool. Extracted subtle information from auxiliary sources to improve the performance of fault localization and program repair.

Stochastic Logic Ltd Dhaka, Bangladesh
July 2015 - August 2016

Quantitative Software Developer

Supervisor : Dr. Arif Dowla
Designed a robust model to estimate Value-at-risk(VAR) from stock market data. Also developed a intelligent model to manage stock portfolio.

Ahsanullah University of Science and TechnologyDhaka, Bangladesh
October 2014 - August 2016

Lecturer

Taught undergraduate course on digital system design and compiler. Supervised undergrad research project.

Kona Software Lab Ltd.Dhaka, Bangladesh
June 2014 - October 2014

Jr. Software Engineer

Designed Proof of Concept of Host Card Emulation System with VISA based payment networks.

Professional skills

Programming Languages

  • Java
  • Python
  • C
  • C++
  • SQL
  • php
  • R
  • MATLAB

Software Tools/Scripting Languages

  • LLVM/Clang
  • Theano
  • Pytorch

Education

Bangladesh University of Engineering and TechnologyMarch 2009 - July 2014

Bachelor Of Science in Computer Science and Engineering