Building Intelligent Systems,
from AI Models to Real-World Products
I design and train AI models, then integrate them into real-world systems. My expertise spans mobile apps, embedded systems, autonomous drones, cloud platforms, and databases — creating intelligent solutions that solve complex problems.
About Me
I love turning ideas into real, working apps. Over the years, I've built skills across mobile development, cloud platforms, and smart technologies. I focus on creating apps that are easy to use and solve real problems.
My Technical Skills
AI & Machine Learning
View DemoAutonomous Systems & Drones
Embedded Systems & IoT
Cloud, Backend & Databases
Mobile Applications (iOS & Android)
Web Development
My Featured Projects
Smart Home Controller
Enables remote home automation with AI-powered security using face recognition and voice commands. Integrates embedded systems (STM32) with mobile apps (Flutter, SwiftUI) and cloud infrastructure (Firebase, AWS) for real-time device control and intelligent safety alerts.
Watch DemoAutonomous Drone System
Demonstrates real-time computer vision for autonomous decision-making. Using PX4, MAVSDK, and OpenCV, the drone detects shapes and autonomously adjusts flight parameters (altitude, navigation patterns) without human intervention — showcasing AI-driven autonomous systems.
Watch DemoY-Shop E-Commerce App
YSHOP is a full-stack intelligent commerce platform unifying E-Commerce, CRM, ERP, and advanced AI. The recommendation engine is built on vector embeddings — every product interaction is encoded into high-dimensional semantic vectors stored in a vector database, enabling millisecond-precision similarity search. The system builds a persistent taste profile per user, and the LLM queries the vector store directly rather than relying on keyword matching — producing recommendations that feel eerily accurate. Collaborative filtering emerges naturally from vector proximity: users with overlapping embedding clusters surface each other's preferences automatically. Also integrates real-time AI moderation, a 24/7 voice assistant, and drone delivery via NVIDIA Jetson.
Tech: Vector DB • Text Embeddings • LLM • Node.js • Python • Swift & SwiftUI • Flutter • Firebase • NVIDIA Jetson • PX4 • MAVSDK • OpenCV • Drone Delivery • AI Moderation • Voice Assistant
View Details Watch AI Demo Watch Drone Delivery
Noota - Voice Translation App
Eliminates language barriers through real-time AI voice cloning and neural translation. Custom Transformer-based AI models (1.5M parameters, 50K training iterations) integrated with production infrastructure (Swift/SwiftUI, Node.js, Firebase) enable seamless multilingual communication for enterprise environments.
Tech: Swift & SwiftUI • Node.js • Python (Flask) • Firebase • Custom Voice Model
Watch Demo
IOTROCOP Marin - Yacht App
Delivers end-to-end IoT automation from embedded firmware to mobile control. Built entire system architecture — hardware integration, backend infrastructure, and native/cross-platform apps (iOS/Android) — enabling real-time remote monitoring, device control, and secure multi-user access for complex autonomous systems.
View Details
Autonomous AI-Powered Industrial Drone System
Solves large-scale building maintenance through autonomous AI-powered drone systems with computer vision and precision navigation.
Tech: Computer Vision • Autonomous Navigation • RTK-GNSS • Multi-Sensor Fusion • Embedded Systems
Watch DemoMinibusCRM – Smart Fleet Management Platform
A scalable fleet management platform designed for transportation companies to monitor and manage minibuses in real time.
Tech: Flutter • Node.js • REST APIs • Google Maps • PostgreSQL • Cloud Deployment
Watch Live DemoDrone In Action
Real-time autonomous flight tests demonstrating manual and autonomous flight control capabilities.
Live Joystick & Throttle Control
Manual Flight & Real-Time Actuation
Indoor System Validation
Manual Takeoff & Stable Hover
Fully Autonomous Takeoff & Hover
From Manual Control to Full Automation
AI in Action
Custom AI models I trained from scratch — demonstrating real-world machine learning in production systems.
No Search Bar. No Filters. Just Talk.
One Voice, Every Language
Real-Time EMG Gesture Recognition — 95% Accuracy
EMG Gesture Recognition — AI & Biomechatronics Lab, YTÜ Yıldız Teknopark
Real-time hand gesture recognition using EMG signals from a Myo Armband to control a prosthetic arm and quadcopter drone. CNN+LSTM model — CNN extracts muscle activation patterns from 8 EMG channels, LSTM captures temporal dynamics. Key breakthrough: fixing a normalization mismatch between training and live inference jumped accuracy from under 5% to over 94%. Final model: ~95% accuracy, <100ms latency, 6 gestures.
Tech: CNN • LSTM • PyTorch • EMG Signal Processing • Myo Armband • Python • Real-Time Inference
YSHOP — Intelligent Logistics & Real-Time Delivery
Advanced delivery system with live map integration, adaptive courier workflows, and real-time synchronization across customer, store, and admin dashboards.
Real-Time Delivery Navigation with Adaptive Workflow & QR Verification
Multi-View Tracking & Live System Sync Across Dashboards
Last-Mile Arrival & Secure Handover with Proximity Verification
How Custom AI Models Work
Speech-to-Text Model (STT)
1.8B parameters • Transformer encoder
LoRA r16/a32 fine-tuning
Trained on ~480h of audio
Inference: ~320ms
Language Model (LLM)
3.4B parameters • 28-layer decoder
GQA + RoPE architecture
LoRA r32/a64 fine-tuning
120K+ commerce conversations
Product Verification (CNN)
ResNet-50 architecture
94.3% accuracy
Trained on 15K+ images
Real-time inference
Status: Under active development. Multi-turn memory, Arabic support, and lower latency optimizations in progress.
NOOTA — Multilingual Voice Communication
Real-time translation platform where speakers maintain their original voice across different languages. Built with custom voice synthesis architecture through extensive optimization iterations.
Swift • SwiftUI
Node.js • Express.js
Python • Flask • GPU
Firebase • GPU Servers
Status: Under active development. Preparing for deployment on high-performance GPU infrastructure for production-ready real-time performance.
Work Experience
AI & Embedded Systems Researcher
YTÜ Yıldız Teknopark · Part-time | Jun 2026 - Present
Istanbul, Turkey · On-site
Joined the YZ & Biomechatronics Research Laboratory at Yıldız Technical University Technopark to design and optimize intelligent AI frameworks bridging machine learning with embedded hardware. Developing deep learning architectures for real-time bio-signal processing and classification. Engineering scalable software to synchronize neural classification models with robotic and autonomous control systems. Implementing feature extraction pipelines for high-accuracy, low-latency applications. Architecting end-to-end intelligent systems for human-machine interfacing.
Tech: Deep Learning • Bio-Signal Processing • Embedded AI • Python • Real-Time Systems • Human-Machine Interfacing
IoT & Machine Learning Researcher / Developer
Al Mutlaq United Company · Part-time | Sep 2024 - Present
Istanbul, Türkiye · Remote
Built real-time physiological stress monitoring system with IoT sensors, automated data pipeline, and AI-powered annotations. Created web-based workflow automation syncing sensor data with video recordings. Integrated vision-based AI for behavioral annotation and dataset validation.
Tech: IoT • Embedded Systems • Python • Data Science • Computer Vision • TensorFlow • Web Development
View Dataset & CodeSoftware Developer (AI, IoT & Autonomous Systems)
IOTROCOP Technology · Full-time | Oct 2024 - Present
- Built smart home systems using ESP32 and STM32 with real-time control via WiFi
- Created mobile apps with Flutter for remote monitoring and control
- Developed face and voice recognition models trained on real data
- Built backend APIs and connected them to Flutter apps
iOS & Flutter Developer (Embedded Systems)
Labcihazlari.com · Part-time | Feb 2024 - Apr 2026
- Built an IoT control system for biological safety cabinets using STM32 and ESP8266
- Created Flutter apps and websites with offline control capability
- Designed full-stack systems using Flask and Node.js
- Built apps for iOS, Android, web, and macOS
iOS & Flutter Developer for Smart Home IoT
OKATAN SAĞLIK ARGE | Sep 2024 - May 2025
- Built iOS apps with SwiftUI and cross-platform apps with Flutter
- Added smart features like face recognition, voice control, and home monitoring
- Integrated RFID/NFC for secure access and cloud database sync
Education
Bachelor of Software Development
Istanbul Aydin University
GPA: 3.5
Languages
Arabic
Native
English
Fluent
Turkish
Intermediate
Get in Touch
Have a project idea or just want to say hi? I'd love to hear from you!
Find Me On
