Open to opportunities

Muhammad Huzaifa Jawad

@muhammadhuzaifajawad

Message

I build explainable, robust machine learning systems for computer vision, time-series, and scientific data.

Kazakhstan

Message

What I'm looking for

I'm open to Machine Learning Engineer, Applied AI Engineer, Computer Vision Engineer, and Research Engineer roles, remote or based in Central Asia, Europe, Asia, or the US.

I've built research-driven machine learning systems at Nazarbayev University, including multimodal human activity recognition that achieved 84.9% ± 2.3% Detection F1 and outperformed a tuned late-fusion baseline.

For climate ML, I reduced CMIP6 mean temperature bias from 4.58 K to −0.09 K using a Conv1D-LSTM and a custom variance-preserving loss. For my MSc thesis, I classified supernova gravitational-wave equation-of-state signals at over 95% accuracy through complementary 1D waveform and 2D scalogram CNN pipelines.

I use Grad-CAM, Integrated Gradients, saliency, and temporal masking to verify that models learn meaningful physical and temporal structure rather than dataset artefacts.

My work spans data preparation, feature engineering, model development, experimentation, cross-validation, interpretability, and evaluation across visual, sensor, time-series, and scientific data. I also strengthened crowd-counting robustness under adverse weather at National Taipei University of Technology by generating synthetic rain with OpenCV.

Experience

Work history, roles, and key accomplishments

Research Assistant - Multimodal Human Activity Recognition

Nazarbayev University

Jun 2026 - Jun 2026 (0 months)

Built an end-to-end multimodal HAR pipeline (IMU + RGB video), reaching 84.9% ± 2.3% Detection F1 under leave-one-subject-out evaluation (10-seed robustness study) — beating a tuned late-fusion baseline by ~4.9 points (86.2% best seed).

Designed and evaluated a learned GatedFusion architecture, improving subject-independent recognition across seven transition activities.

Machine Learning Deep Learning Computer Vision

Research Assistant - Climate ML / CMIP6 Bias Correction

Nazarbayev University

Aug 2025 - Dec 2025 (4 months)

Cut CMIP6 mean temperature bias from 4.58 K to −0.09 K (~37% RMSE reduction) with a Conv1D-LSTM model using a 14-day input window.

Prevented output collapse via a custom variance-preserving loss (MAE + variance penalty), validated with Mann-Kendall trend and seasonal analysis.

Benchmarked against CycleGAN (collapse-prone on 1D temporal data) and statistical baselines, establishing CNN-LSTM as

Research Skills Machine Learning Deep Learning Computer Vision

Research Assistant - Scientific ML / XEOS-NET (MSc thesis)

Nazarbayev University

May 2025 - Aug 2025 (3 months)

Classified the nuclear equation-of-state from core-collapse supernova gravitational-wave signals at >95% accuracy using a dual-pipeline CNN (1D CNN on raw waveforms; 2D CNN on CWT scalograms).

Verified physical validity with Grad-CAM, Integrated Gradients and temporal masking, showing stiff EOS depend on the bounce phase and soft EOS on ring-down.

Machine Learning Computer Vision

Research Trainee

National Taipei University of Technology

Dec 2023 - May 2024 (5 months)

Improved crowd-counting robustness under adverse weather by generating synthetic rain with OpenCV to expand the training distribution.

Evaluated performance across visibility levels, identifying reliability limits under distribution shift for outdoor computer-vision systems.

Computer Vision PyTorch Deep Learning Machine Learning

Research Assistant

SID - Secured IoT Devices Lab - UET Peshawar

Jan 2023 - Nov 2023 (10 months)

Conducted research in computer vision and deep learning, focusing on IoT security applications.

Computer Vision Research Skills PyTorch

Machine Learning Engineer

Technocolabs Softwares

Oct 2022 - Dec 2022 (2 months)

Worked on data manipulation, feature engineering, and machine learning model development.

Data Manipulation Feature Engineering Data Cleaning Machine Learning Data Analysis Data Visualization

Core Team Member of Developer Student Club

Google Developers

Sep 2020 - Jul 2021 (10 months)

Contributed to organizing and leading developer community events and workshops.

Core Team Member of Developer Student Club | DSC UET Peshawar

Google Developers

Sep 2020 - Jul 2021 (10 months)

Worked as Core Team Member of Developer Student Club | DSC UET Peshawar at Google Developers.

Career-Prep Fellow

Amal Academy

Mar 2021 - Jun 2021 (3 months)

– Communication: Completed a competitive written application and interview process to be selected from over 5000+ applicants for intensive 3-month Fellowship funded by Stanford University
– Skills development: Invested 150 hours in order to develop business skills (e.g., communication, leadership, problem solving, teamwork, etc.) that will help me make a deeper impact on the job