Hi, my name is
Faizan Aziz.
I extract insights from data.
I'm a data scientist and machine learning engineer focused on deep learning, predictive analytics, computer vision, and large-scale data processing. I develop solutions that uncover patterns and drive data-driven decisions.
Get In Touch!About Me
Hello! My name is Faizan and I enjoy building intelligent systems. I'm currently pursuing my Bachelors in Data Science at FAST NUCES, where I've been on the Dean's List of Honors multiple times. My passion lies in computer vision, deep learning, and creating production-ready AI systems.
Fast-forward to today, and I've had the privilege of working as an AI Engineer at DMLytics, where I architected and implemented production-ready preprocessing pipelines for drone imagery and automated flight planning systems.
Here are a few technologies I've been working with recently:
- Python
- PyTorch
- TensorFlow
- Computer Vision
- Deep Learning
- OpenCV
- Scikit-learn
- SQL
Where I've Worked
AI Engineer @ DMLytics
Feb 2025 - June 2025
- Architected and implemented a production-ready orthorectification preprocessing pipeline for drone imagery, handling the full system lifecycle
- Processed 1,400+ high-resolution aerial images (~25MB per image), extracting EXIF and flight metadata to normalize orientation using pitch, roll, and yaw (PRY) values
- Applied geometric rotations and spatial normalization to standardize imagery prior to georeferencing for consistent spatial alignment across large-scale agricultural datasets
- Optimized storage and downstream processing by converting corrected imagery to JP2 format while preserving spatial and radiometric fidelity
- Designed and co-developed a KML/KMZ-based flight plan generation system in a 4-member team to create structured aerial survey paths
- Implemented multi-pass flight planning logic combining high-altitude coverage with low-altitude reverse passes for detailed crop health analysis
- Automated generation of finalized KMZ flight plans compatible with drone mission software and geospatial visualization platforms
Some Things I've Built
-
Featured Project
Vigilon - AI-Powered Smart Surveillance System
Built a CCTV surveillance system using Computer Vision for weapon detection, intrusion detection, and behavioral anomaly analysis. Implemented YOLO-based object detection and adaptive anomaly detection. Designed an end-to-end pipeline for video ingestion, threat classification, alerting, and backend integration.
- Python
- YOLO
- Computer Vision
- OpenCV
- Deep Learning
-
Featured Project
Scalable Streaming Join Engine using MeshJoin
Implemented the MeshJoin algorithm to join high-volume transaction streams with disk-resident dimension tables. Designed memory-partitioned buffers and queue-based stream processing for incremental joins. Integrated duplicate detection and disk buffering to simulate streaming ETL and real-time data warehousing.
- Python
- Streaming
- Data Processing
- ETL
- Algorithms
-
Featured Project
Multi-Task Facial Emotion Analysis
Built a facial emotion analysis pipeline using Computer Vision and Multi-Task Deep Learning. Applied transfer learning with VGG16 and ResNet50 in TensorFlow and Keras. Performed data preprocessing, facial landmark normalization, and evaluated classification and regression performance.
- Python
- TensorFlow
- Keras
- VGG16
- ResNet50
- Computer Vision
Other Noteworthy Projects
view the archiveWhat's Next?
Get In Touch
I'm always interested in hearing about new opportunities in AI engineering, machine learning, and data science. Whether you're hiring, have an interesting project, or just want to connect, feel free to reach out!
Say Hello