Sallar A. Qazi

Research Engineer | Scientific Computing, Machine Learning & High-Performance Computing

qazisallar@gmail.com | London, United Kingdom

About Me

Research Engineer specializing in scientific computing, optimization, machine learning, and high-performance computing; experienced in developing numerical solvers, deep learning models, and AI-driven simulation frameworks. Passionate about tackling challenging problems and continuously expanding technical expertise.

Experience

Early Stage Researcher | Imperial College London (2021–2025)
  • Developed transformer-based image segmentation models for medical imaging.
  • Designed multiscale FE/BE-based algorithms for knee joint simulations using Python/FEniCS.
  • Applied Physics-Informed Neural Networks (PINNs) to predict micro-contact wear.
  • Presented findings at international conferences; authored peer-reviewed papers.
Engineer (Various Roles) | Pakistan Petroleum Limited (2016–2019)
  • Led mechanical teams for commissioning wellhead compression units.
  • Conducted predictive analysis and maintenance on critical machinery.
  • Planned and executed major overhauls for plant engines and turbines.

Education

  • Ph.D. Computational Sciences with focus on Biomechanics and BioTribology, Imperial College London (2021–2025)
  • Erasmus Joint Masters in Tribology of Surfaces and Interfaces , University of Leeds, University of Ljubljana, Luleå University of Technology (2019–2021)
  • BE Mechanical Engineering, National University of Sciences and Technology (2012–2016)

Awards & Honors

  • Marie Skłodowska-Curie Fellowship (2021)
  • Erasmus Mundus Scholarship (2019)
  • Field Performance Award – Pakistan Petroleum Limited (2018)
  • Best Employee Award – Pakistan Petroleum Limited (2017)
  • TEDx Speaker (2016)
  • Most Effective Strategy Award – Formula Student UK (2015)

Skills

  • Programming & Tools: Python, C++, Julia, R, NumPy, SciPy, PyTorch, MATLAB, JAX
  • Data Visualization: pandas, matplotlib, pyspark, PowerBI, Tableau
  • HPC: OpenMP, MPI, CUDA
  • Scientific Computing: FEM, BEM, Optimization, Statistics

Deep Learning Projects

  • Transformer-based Image Segmentation – Medical imaging segmentation models for joint outlines.
  • Physics-Informed Neural Networks (PINNs) – Predictive models for surface wear.
  • Knowledge Distillation Pipeline – Efficient student models with reduced parameters.