Looking for a job

Ahmad Ramiz Qeyam

@ahmadramizqeyam

Machine learning engineer who denoised bus GPS trajectories at InnoForce Group and built a loan-risk model with 0.738 ROC-AUC.

Kazakhstan

Message

At InnoForce Group, I worked on denoising bus GPS trajectories for public transport monitoring. I applied Exponential Moving Average, Savitzky-Golay, and Kalman filtering, and built a preprocessing and evaluation workflow for GPS data.

For my LendingClub Loan Risk Prediction project, I built an end-to-end machine learning pipeline on approximately 2.26 million loan records and compared classification models. I tuned the classification threshold and achieved 0.738 ROC-AUC and 0.420 PR-AUC on the test set.

Experience

Work history, roles, and key accomplishments

Machine Learning Engineering Intern

InnoForce Group

Jan 2026 - Mar 2026 (2 months)

Worked on bus GPS trajectory denoising using outlier rejection and smoothing techniques for public transport monitoring. Applied EMA, Savitzky-Golay, and Kalman filtering to reduce GPS noise and improve trajectory quality.

Python Geospatial Data Processing EMA Savitzky Golay Filters Kalman Filter data preprocessing Trajectory Analysis Anomaly Detection

AI Research Intern

Al-Farabi Kazakh National University

Jun 2024 - Aug 2024 (2 months)

Conducted research on machine learning algorithms for classification and predictive analytics applications. Analyzed and cleaned large datasets, improving data quality and supporting more reliable model outcomes.

Machine Learning Data Cleaning Data Analysis Research Technical Documentation