Industry and applied research

  1. Bid Optimization Analyst
    Louis W Bray Construction Limited
    • Automated the structuring of more than 4,000 unstructured tender sheets, removing a recurring manual review cycle.
    • Developed a bid prediction tool combining custom parsing with regression models.
    • Produced market and competitor analyses that informed pricing and win strategy.
  2. Data Scientist Intern
    Louis W Bray Construction Limited, MITACS Business Strategy Internship
    • Worked with the estimating and finance teams on an end-to-end data science programme.
    • Automated ingestion and analysis of bid documents, reducing manual review time while surfacing margin and cost insights.
    • Built monitoring dashboards with predictive signals so that leadership could track project risk, win probability, and cash exposure in near real time.
  3. Machine Learning Engineer
    ACI Limited, Bangladesh
    • Built a social media analytics system measuring sentiment around product launches, associated with a 15 percent increase in brand engagement.
    • Developed an RFM-based customer segmentation model that raised conversion rates by 18 percent.
    • Engineered a recommendation system using purchase patterns, growing cross-segment sales by 12 percent.
    • Designed predictive models to identify emerging clinicians from FourP data, supporting a 10 percent increase in new client acquisition.
    • Created credit risk dashboards that contributed to a 7 percent reduction in default rates.
    • Delivered business intelligence analyses across several operating units.
  4. Research Student
    MyMedicalHUB
    • Built models analysing risk factors for musculoskeletal pain and identifying symptom patterns.
    • Developed facial expression, age, and sex recognition from webcam input.
    • Contributed to the design of an automated symptom checker.

University appointments

  1. Integrated Project Delivery Centre Intern Current
    Arthur B. McDonald Canadian Astroparticle Physics Research Institute

    Queen's Doctoral Internship in University Administration.

  2. Ph.D. Graduate Committee Officer
    Queen's Graduate Computing Society
  3. Project Assistant
    Career Services, Queen's University
  4. Officer, Promotions
    Queen's Bangladeshi Students' Association

Teaching appointments are listed on the teaching page.

Projects

  • Queries a Google Scholar profile and produces a static publications page, separating journal and conference entries, emphasising the author's own name, and attaching a copyable BibTeX record to each item. It generates the research page on this site.

    Open source 2025
  • A Python library that reports statistics for pull requests, with plotting utilities suitable for dashboards and notebooks.

    Open source published on PyPI
  • Automates fifteen common exploratory analysis and cleaning routines in order to shorten the interval between receiving a dataset and understanding it.

    Open source published on PyPI
  • Privacy-Preserving Machine Learning with Homomorphic Encryption

    An examination of homomorphic encryption schemes for protecting sensitive machine learning workflows, validated on confidential datasets using linear regression.

    Python Jupyter
  • Latency of Time to First Response in GitHub Pull Requests

    An analysis of response latency in pull request workflows. The study developed here became the basis of the MSR 2023 paper.

    Course project CISC 834
  • Implementations of Neural Network Algorithms

    Laboratory implementations covering nearest neighbour classification, perceptron variants, Kohonen self-organising maps, and Hopfield networks.

    Python Jupyter

Research tools and replication packages are listed on the tools page.

Undergraduate thesis

Prediction of Clinical Risk Factors of Diabetes Using Multiple Machine Learning Techniques Resolving Class Imbalance. Part of this thesis was published at the 23rd International Conference on Computer and Information Technology in 2020, supervised by Asst. Prof. Barshon Sen. Paper, presentation.

Kaggle

Kaggle Expert in four categories: Competitions (ranked 844 of 162,151), Notebooks (262 of 140,050), Discussion (283 of 162,435), and Datasets (124 of 25,284). Individual competition results appear on the awards page.