AI & Software Engineer • Mobile Development Lead • Co-founder of sysHuman
30+ years bridging scientific computing, signal processing, and production AI — from Istanbul Technical University research to LLM- and computer-vision-powered products shipping in Canada today.
Software engineer with 30+ years of hands-on delivery — architecting, shipping, and leading production systems across AI, mobile, backend, and embedded. Co-founder and Technical Director of sysHuman Technologies Inc. (Vancouver), shipping LLM and computer-vision powered iOS, Android, and web products end-to-end: model serving on GPU, REST/PHP backends, native mobile clients, StoreKit 2 and Google Play Billing. Mobile team lead at 4Refuel — enterprise Android (MVVM, RxJava2, Retrofit, Room) with a proprietary BLE/LCR protocol to IoT fuel-dispensing hardware. Formerly Android developer on Intel's Vaunt smart glasses and backend AI developer at Prizm Media (SVM activity recognition in production). Granted patent in AI-based sensing.
Method and Apparatus for Detecting Color — ML-based real-time color classification via light sensor hardware. Granted.
Ph.D. coursework & M.Sc. — Geophysical Engineering, İTÜ (Istanbul). B.Sc. — Yıldız University (Istanbul). iOS Development — BCIT (Vancouver).
North Vancouver, British Columbia, Canada.
Vancouver-based multinational AI company delivering LLM + computer-vision products across iOS, Android, and web. Led founding team across Canada and Turkey on product, mobile, and operations. Resolved multiple App Store review rounds; implemented StoreKit 2 (iOS) and Google Play Billing (Android). Built an automated AI content pipeline on Ubuntu 24.04 (Claude Code + cron) publishing to Instagram and TikTok.
Leading the mobile engineering team behind the enterprise Android app for fleet fuel management across Canada. MVVM, RxJava2, Retrofit, Room. Designed the proprietary LCR/LCP hardware protocol with CRC16 verification, and the real-time BLE bridge between Android devices and IoT fuel-dispensing hardware in the field. Architecture, code review, and mentorship for junior developers.
Production LLM + computer-vision platform — 4-in-1 beauty analysis (skin, hair, makeup, outfit). Live on iOS, Android, and web. Token-based IAP. Studio section with SDXL inpainting + IP-Adapter FaceID on RTX GPU.
LLM · Computer Vision · StoreKit 2 · Google Play Billing
mirror.syshuman.comB2B AI skin-analysis SaaS for beauty salons. Professional PDF reports, multi-branch management, pay-per-token model, Groq and Ollama backends, multilingual (Turkish/English).
PHP · MySQL · Groq · Ollama
salon.syshuman.comResearch platform applying machine-learning models to seismic signal data — drawing directly on my geophysics background at İTÜ.
ML · Seismic Signal Processing
quake.syshuman.comSelf-hosted beauty-transformation API: SDXL Inpainting + BiSeNet segmentation + BeautyGAN + IP-Adapter FaceID, running on an RTX 4070Ti server.
SDXL · BiSeNet · BeautyGAN · CUDA
image.syshuman.comFull AI-generated Turkish-language science documentary in production using Veo 3 generative video — scripted and narrated by Kadir.
Veo 3 · Generative Video · Turkish Narration
YouTube ChannelGranted patent — "Method and Apparatus for Detecting Color." ML-based real-time color classification using custom light-sensor hardware.
Patent · Embedded ML · Light Sensor
Android + BLE developer on Intel's publicly announced smart-glasses project (The Verge, Feb 2018). Gesture and sensor-interaction layers for wearable AR hardware.
Android · BLE GATT · MVP · Dagger
Real-time FFT audio signal visualisation. 3,000+ downloads on Google Play.
FFT · Real-Time DSP · Android
R&D project with UBC Mechatronics Division — real-time vibration/chatter detection using signal processing.
Signal Processing · Vibration Analysis
youtube.com/@KadirErturk — 7+ years of deep, mathematically-grounded AI and science education in Turkish (and some English). Topics go well beyond introductory material.
Transformer architecture & self-attention (full mathematical build-up), query/key/values, backpropagation from scratch, neural network fundamentals, feature vectors, embedding layers.
Matrices, eigen vectors & eigen values, differential (difference) equations, differential calculus foundations.
LSTM stock-price prediction (7K views), Q-Learning with example (2.5K), AI-based FX prediction (4.3K), "Creating life with AI" series.
Apoptosis — why your cells commit suicide, Architects of Death, writing code on DNA, sensor hardware demos.
Vancouver, Canada. AI products across iOS, Android, and web.
Surrey, BC. Led enterprise Android app + BLE/IoT hardware integration for fleet fuel management.
Vancouver. Full-stack — web, Android, iOS, REST APIs.
Vancouver. Intel Vaunt smart glasses — Android + BLE. Internal ML presentations on activity recognition and SVM-based signal classification.
Vancouver. Trained SVM for real-time human activity recognition (deployed in KudoLife). Neural networks for user behaviour classification.
North Vancouver. Published TraceMe (App Store + Android Market), Fourier Transform, ChatterPro (UBC collaboration).
Istanbul. Production automation software for Turkish leather industry — became industry standard with 120+ worldwide customers.
Istanbul. ANSI C/C++ barcode applications for handheld computers — Coca-Cola, Philips, Levi's.
Engineering faculty. Scientific software in Fortran and C/C++; peer-reviewed publications.
Hands-on, step-by-step: from data to deployed model. Practical how-to focus.
Mathematical foundations — attention, self-attention, positional encoding, encoder-decoder. Rigorous theoretical depth.
Java-centric software engineering and object-oriented design.
iOS and Android — native and cross-platform. From architecture to App Store / Play deployment.
ESP32, BLE, light sensors, RF circuit design (ADF4351, NanoVNA V2).
Fourier/Laplace/Z transforms, 2D/3D image processing, real-time DSP.
Large Language Models, Groq API, Claude API (Anthropic), Ollama, LangChain, Prompt Engineering, RAG pipelines, LLMOps.
Neural Networks, SVM, LSTM, deep learning, Transformer architecture, activity recognition, computer vision, FaceID.
NVIDIA CUDA, RTX 3060 / 4070Ti, GPU-accelerated inference, Ubuntu 22/24 server administration.
Android — MVVM, RxJava2, Retrofit, Room, BLE/GATT, StoreKit 2, Google Play Billing. iOS — Swift, Objective-C, Xcode.
Java, Python, PHP, REST API design, SmartGWT, Yii, Node.js, Apache. HTML/CSS/JS, jQuery, AJAX, React (Tailwind).
MySQL, SQLite, Room, Firebird, Interbase, PostgreSQL. DDL/DML, stored procedures, triggers, replication.
ESP32, BLE (GATT), RF circuit design (ADF4351, NanoVNA V2), light sensors, barcode (ANSI C), Fortran, C/C++.
Fourier/Laplace/Z transforms, 2D/3D image processing. Linux (Ubuntu), Git, Docker, Claude Code, cron, Android Studio, Xcode.