Pooya Nabavi
Camera Engineer β’ Deep Learning Expert β’ Tech Lead
About me
I'm Pooya Nabavi, Ph.D., a Camera Design Engineer (Tech Lead) at Apple. My work sits at the intersection of advanced machine learning, full-stack software, and next-generation camera systems. With over a decade of experience spanning Apple and optical wireless technologies, I focus on creating high-impact, integrated solutions that push the boundaries of imaging, computer vision, and intelligent systems.
Current Focus
- AI-driven camera and computer vision architectures
- Neural networks for automated inspection and failure prediction
- Full-stack + ML pipelines for intelligent imaging platforms
- Real-time machine learning for embedded vision systems
- Adaptive detection in optical wireless communication
"I cast light on complexity, turning the opaque into insight, execution and efficiency."
Selected Projects
Machine Learning & Signal Processing
-
Customer Clustering for Behavioral Segmentation in Retail
Applied K-Means, Gaussian Mixture Models, PCA, and t-SNE to uncover psychographic customer groups using personality, demographic, and transactional data. Enabled targeted marketing and resource optimization for a major retail client.
-
Supervised Classification Pipeline for Lead Scoring β ExtraaLearn
Built interpretable, high-recall classification models using Decision Trees and Random Forests with categorical encoding and hyperparameter tuning. Enhanced lead targeting efficiency for a growing ed-tech firm.
-
Adaptive Equalization for Underwater VLC
Designed symbol-level adaptive algorithms for underwater visible light communication. Used recursive estimation and multi-symbol detection to combat channel distortion.
Optics, Embedded, & RF Systems
-
10β f/5.5 Achromatic Telephoto Lens Design
Designed diffraction-limited, high-magnification telephoto optics using ZEMAX. Achieved chromatic and spherical aberration correction through symmetric layout.
-
40x Microscope Objective Design
Engineered achromatic 40x, 0.6 NA objective for F/e/C spectral lines using ZEMAX. Compact, high-resolution design optimized for academic and industrial microscopy.
-
BER Analysis of Non-Stationary Optical Fields
Modeled BER for OOC-CDMA infrared systems using saddle-point approximation in MATLAB. Focused on high-speed non-stationary optical channels.
-
JPEG Huffman Decoder on FPGA
Implemented hardware-accelerated Huffman decoder for JPEG compression using Verilog on FPGA. Prioritized pipeline efficiency and hardware cost optimization.
-
UAV Pulse Radar for Smuggling Detection
Designed radar for airborne smuggling detection in low-visibility desert zones. Recognized with 1st-place award for innovation and feasibility.
-
Broadband Microwave Matching Network
Simulated impedance matching networks (Stub, Binomial, Chebyshev, Tapped-L) using ADS and Smith Chart. Built MATLAB GUI for rapid design automation and optimization.
Education
-
Ph.D., Electrical Engineering
University of Central Florida, Aug 2017 β May 2022
Thesis: Multi-Element Mobile Optical Wireless Communication Networks -
M.Sc., Optics & Photonics
CREOL, University of Central Florida, Aug 2019 β Dec 2020 -
M.Sc., Electrical Engineering
Sharif University of Technology, Aug 2015 β Aug 2017
-
B.Sc., Electrical Engineering
K. N. Toosi University of Technology, Aug 2011 β Aug 2015
Technical Skills
- Scripting: JMP (JSL), MATLAB, Apple Script, Bash
- Languages: Python, JavaScript, C#, SQL
- Frameworks: Django, Angular, HTMx, Bootstrap, HTML, CSS
- ML Libraries: NumPy, Pandas, SciPy, Scikit-learn, PyTorch, TensorFlow