Straight to the Solution
Mobile & On-Device AI Engineer with a strong foundation in Computer Engineering & Biomedical Engineering (Accredited by MÜDEK) at Ankara University. I build Android apps that run AI on-device with zero cloud latency, zero data leaks, and no compromises. My work spans scientific instrumentation (NFC potentiostats), on-device LLM inference (Gemma 3), and productivity tools, all published on Google Play under SensiStudio. I care about shipping real products fast and making complex hardware feel natural on a phone.
Work Experience
AI Intern (Microsoft AI Innovators SummerSchool 2026 Program)
Selected for the Microsoft AI Innovators program to independently engineer a high-performance, privacy-first AI solution.
- Architecting a secure, enterprise-grade AI system using Foundry, specifically designed to operate 100% local and offline.
- Responsible for handling the entire development lifecycle and optimizing local model deployment for zero-connectivity environments.
- Refining the architecture through weekly engineering iterations and benchmark evaluations.
R&D Software Developer (Working Student)
Owned the full development lifecycle of the Sensitify Go Android app, from architecture to Google Play Store release.
- Shipped a mobile potentiostat-control app supporting 9 electrochemical techniques (CV, DPV, SWV, EIS, CA, LSV, NPV, OCP, Potentiometry) with real-time plotting.
- Built USB-host device communication for ECA/MCA hardware and data export to .xlsx / .m formats.
- Fine-tuned an on-device LLM for voice-controlled electrochemistry guidance, fully offline, zero data collection.
Engineering Intern
Improved the Sensitify Studio desktop app and the company's official web presence.
- Rebuilt UI components in C++ / Qt & QML for a smoother instrument-control experience.
- Designed and developed the company's official website, strengthening brand identity.
Engineering Intern
Maintained, repaired, and calibrated medical devices across multiple hospital departments as a mandatory Biomedical Engineering clinical internship.
Core Skills
On-Device AI & Local LLMs
Gemma 3, MediaPipe LLM Inference, Microsoft Foundry, Local RAG, CMSIS-NN, Edge Inference & Quantization
Mobile & Hardware Interfacing
Jetpack Compose, Kotlin Coroutines/Flow, USB Host (CDC/FTDI), ST25DV NFC Mailbox, Real-Time Plotting Engine
Embedded & Hardware Design
STM32 (ARM Cortex-M), ESP32-C3, 4-Layer PCB Design (KiCad), NFC Energy Harvesting, Fusion 360 CAD
Systems & Native Tooling
Modern C/C++ (C++17/20), Python, Qt 6 / QML Desktop, Git CI/CD, Google Play Release Lifecycle
Tech Stack
Projects & Published Apps
Foundry Local RAG – Privacy-First Offline AI
Selected for the Microsoft AI Innovators program to independently engineer a high-performance, privacy-first AI solution. Architecting a secure, enterprise-grade AI system using Foundry, specifically designed to operate 100% local and offline with zero cloud dependency.
Sensitify GO – EIS On The Go
Professional electrochemical analysis app for MCA devices. Controls potentiostat hardware via USB, supports 9 measurement techniques, real-time plotting, and data export in .xlsx and .m formats. Zero data collection, fully offline.
NFC Potentiostat – Battery-Free Electrochemical Platform
A fully wireless, battery-free potentiostat powered entirely by NFC energy harvesting from a smartphone. Custom 4-layer PCB (STM32U545 + ST25DV), 3D-printed magnetic enclosure, CNC-milled 13.56 MHz antennas, and a companion Android app with real-time plotting and an on-device Gemma 3 1B AI agent for voice-controlled analysis. TÜBİTAK 2209-B approved.
Clinical No-Contact IR Thermometer
DIY clinical-grade non-contact infrared thermometer designed according to international EN ISO 80601-2-56:2017 medical standards. Features custom PCB hardware (ESP32-C3 Mini, MLX90614 IR sensor, TP4056 USB-C Li-Po charging, SSD1306 OLED display, active buzzer), Clean Code C++ firmware, ergonomic 3D-printed enclosure, and complete laboratory verification documentation.
Tranquile – Meditations by Local AI
Generates personalized guided meditations entirely on-device with Gemma 3, 100% offline, no account, nothing leaves the phone. Describe what's on your mind and the AI writes a unique session in seconds with 5 narrator voices (TTS), real-time sentence highlighting, save & replay, and streak tracking.
LibreChat | Run AI Locally!
100% offline on-device LLM chat powered by Google's Gemma 3. No accounts, no tracking, and chats deleted when app closes. Voice assistance with LibreChat Live Mode. Privacy-first AI on your phone.
Local AI Benchmark
Benchmark your device's on-device LLM performance. Measures tokens/sec, CPU clock speed, and RAM usage in real time with dynamic charts. Scoring system ranks devices from Entry-Level to Flagship Class. Built with Kotlin & Jetpack Compose.
Nuclear Routines: Life-Changer
Habit and routine tracker that helps determine where, when, and how habits occur. Personalized tracking with reminders, achievements, and Turkish language support.
ReadFacile: Read Easier!
Speed reading app based on scientific research: lighter-colored word suffixes let the brain predict and read faster. Export as PDF or DOCX. Designed for people with learning difficulties.
Academic Education
B.Sc. in Computer Engineering
Software engineering, algorithms, data structures, OOP, database systems, deep learning. Selected for Microsoft AI Innovators Summer School 2026 (Edge-AI & On-Device AI), instruction language is 100% English.
B.Sc. in Biomedical Engineering
Biomedical instrumentation, electrochemistry, HCI usability studies, medical device technologies, instruction language is 100% English.