Atlanta, GA | (404) 433-4290 | zarrabinikou@gmail.com
LinkedIn: nikouzarrabi |
GitHub: nkzarrabi
Applied AI/ML Engineer with MS in Computer Science (Interactive Intelligence) specializing in machine learning model training, image-based computer vision, and production systems. Experience with supervised ML model development on structured and spatial datasets, implementing probability map generation, and deploying inference pipelines via REST APIs. Background in document computer vision (OCR), visual reasoning tasks, and building scalable production software with Python, Django, and cloud platforms.
Applied ML & Computer Vision Engineer (Football Analytics) Contractor | 2023 – 2024 - Trained supervised ML models on structured and spatial football event data (Opta, Wyscout format), performing feature engineering and parameter tuning for predictive analytics - Built regression models for player performance metrics and market value estimation, generating out-of-sample predictions integrated into analytics dashboards - Reconstructed pixel-aligned probability maps from event-level data to replicate Expected Goals (xG) visualizations and heatmap analytics - Validated ML model outputs against image-space ground truth, ensuring spatial accuracy of probability surface generation - Deployed inference outputs via REST APIs for integration into downstream analytics systems
iOS Developer Swimming Cars | Jan 2025 – Present - Built production SwiftUI iOS app for carpool coordination serving 50+ users, integrating Firebase for real-time data synchronization and cloud persistence - Implemented geolocation tracking system using Apple Maps SDK with event-driven architecture for state management - Developing REST API integration layer for model inference endpoints and ML service integration - Deployed cloud-native mobile application with zero crashes through comprehensive testing and CI/CD automation
Front Desk Associate Stone Summit Atlanta | 2022 – 2023 - Managed front desk operations in a high-traffic climbing gym environment - Handled memberships, check-ins, and customer interactions - Maintained efficient workflow during peak hours while ensuring safety compliance
DevOps Engineer / E-commerce Tech Lead Keivan Woven Arts | May 2024 – Jan 2025 - Built and deployed automated infrastructure for product listings across 3 platforms (Shopify, Chairish, 1stdibs), reducing manual processing time by 80% - Implemented CI/CD pipelines using GitHub Actions with automated testing and validation, reducing deployment errors by 95% - Integrated AWS S3 with Django backend for scalable image storage serving 1,000+ SKUs, improving performance by 60% - Developed Python automation scripts for cross-platform inventory synchronization and data validation
Data Automation Intern Porsche Cars North America | Jun 2018 – Nov 2018 - Developed OCR-based document processing pipeline for invoice automation, extracting structured data from scanned invoices with 99% accuracy across 500+ documents per week - Implemented PDF parsing system using OCR technology to digitize paper-based workflows, reducing manual processing time by 75% - Built production computer vision pipeline integrating OCR outputs into automated workflows, eliminating 10,000+ pages annually during critical business transition
AI-Powered Employment Assessment Platform | GitHub
- Trained supervised ML models for candidate assessment using
TensorFlow, performing feature engineering and model tuning for
role-specific evaluations - Developed predictive scoring system with
bias-free evaluation criteria, integrating trained models into
production inference pipeline - Built React frontend and Django backend
REST API for serving model predictions and supporting real-time
assessments
Visual Reasoning AI System | Georgia Tech
Coursework
- Implemented visual reasoning algorithms solving Raven’s Progressive
Matrices using PIL for image processing - Developed pattern recognition
and spatial relationship extraction directly from image data -
Demonstrated symbolic AI approach to visual reasoning, identifying
geometric transformations and logical patterns without deep learning
Django Health Tracker (Raspberry Pi) | GitHub
- Full-stack health monitoring application with Django backend and
JavaScript frontend, deployed on Raspberry Pi - Implemented real-time
data visualization and threshold-based alerting for personal health
metrics tracking
Package Sorting Algorithm | GitHub
- Developed classification logic for robotic automation system with
comprehensive test coverage - Implemented rule-based algorithm for
package categorization based on dimensional and weight constraints
MS in Computer Science — Interactive
Intelligence
Georgia Institute of Technology, August 2023
Coursework: Artificial Intelligence, Knowledge-Based AI (Visual
Reasoning), Machine Learning, Robotics: AI Techniques (Perception,
Spatial Reasoning), Software Architecture & Analysis
BS in Mathematics
University of Georgia, May 2018
Member Common Climbing (CTR Non-Profit Rock Climbing Club) - Participated in community-driven climbing events focused on accessibility and inclusion
Member Southeastern Climbers Coalition (SCC) - Contributed to trail days supporting maintenance and access for climbing properties - Participated in stewardship efforts to preserve outdoor climbing areas