CS student bridging embedded systems, machine learning, and full-stack web. From ESP32-powered kitchen appliances to AI-enabled smart glasses, I ship things that work in the real world.
Building at the intersection of hardware and intelligence. Work spans embedded systems, machine learning, and full-stack software, from sensor-driven prototypes to computer vision models and deployed web platforms.
Comfortable across the stack: circuits, firmware, backend systems, and user-facing applications. Most projects begin with curiosity and evolve into deep technical exploration.
Studying Computer Science while building independently for clients, research, and personal exploration. Long term: creating technology that matters at scale.
Group project (team of 4). I owned the full software stack: Raspberry Pi integration, LiDAR depth processing, and Gemini API pipeline for real-time scene understanding. Multimodal AI running at the physical edge.
Automated a regular rice cooker using ESP32 and a motorized mechanism, enabling remote scheduling and control. Pure hardware hacking without fancy sensors.
AI-powered receipt scanning platform for restaurants. Gemini 2.5 Flash reads any food delivery receipt in about 2 seconds, extracting customer name, phone, and order total. Automates WhatsApp follow-ups and feeds a multi-location analytics dashboard. 98% accuracy, zero manual entry.
Trained a CNN from scratch using PyTorch to classify cats and dogs. Deployed as a live web app where users can upload images and get instant predictions.
Client project: a mentorship platform guiding aspiring pilots through their aviation career. Built and deployed end-to-end with real users actively using it.
Bilingual (English/Farsi) multi-page guide covering history, brewing techniques, and tea varieties. Demonstrates i18n, multi-page routing, and content design.
Interactive educational tool teaching Karnaugh map simplification for digital systems design. Makes a traditionally dry topic approachable and visual.
Web-based data gathering tool for ADHD research. Collects user responses and pipelines results directly into Google Sheets for analysis.
iOS app for keeping score in Shelem, a popular Persian card game. Supports voice entry in English and Farsi, tracks bids and results across rounds, and shows live running totals. Built for iOS 17+ with full Persian right-to-left support. Available free on the App Store.
Microcontroller expertise with IoT development, motor control, and hardware automation for real-world embedded systems.
Edge computing and Linux-based development for real-time applications, GPIO interfacing, and AI deployment on constrained hardware.
Depth sensing and spatial data acquisition for computer vision and robotics applications.
Deep learning and CNN training from scratch for computer vision tasks and model deployment.
Multimodal LLM integration for real-time AI inference on edge devices with image and text processing.
Full-stack React framework for building scalable web applications with modern tooling and deployment pipelines.
Primary language for ML, scripting, embedded systems, and data pipeline development across projects.
Low-level firmware and embedded systems programming for performance-critical microcontroller applications.
Cloud deployment and CI/CD pipelines for seamless production hosting and edge function deployment.
Backend integration for data pipelines, form submissions, and real-time data collection workflows.
Image processing with OpenCV and neural networks for object detection and scene understanding tasks.