name: Mehmet GΓΌmΓΌΕ
location: Istanbul, Turkey πΉπ·
current_roles:
- AI Trainer @ Invisible Technologies
- Embedded SW Team Lead @ UKET (Volunteer)
philosophy:
- "Physics first, code second"
- "Simulate before you build"
- "Measure everything, assume nothing"
focus_areas:
- Aerospace & Defense Software
- AI/ML & LLM Training (RLHF)
- Real-time Physics Simulations
- Embedded Systems & Robotics
- 3D Visualization & WebGL
fun_fact: "I build rockets in software before they fly in reality π"| Role | Organization | Period | Highlights |
|---|---|---|---|
| π€ AI Trainer | Invisible Technologies | 2025 - Now | LLM validation, RLHF, Turkish NLP |
| π Founder & Developer | Bilir.app | 2024 - Now | Flutter, OOP design, 6-person team lead |
| βοΈ Embedded SW Lead (Volunteer) | UKET | 2023 - Now | Rover systems, sensor fusion, C/C++ |
| π¬ Software Intern | TΓBΔ°TAK MAM | 2023 | DSP, VNA data, MATLAB/Python |
Rocket Engine & Flight Simulation Platform
π¬ NASA CEA methodology with Gordon-McBride equilibrium solver
βοΈ Gibbs free energy minimization for combustion analysis
π 6-DOF flight simulation with quaternion-based orientation
π Monte Carlo dispersion analysis for landing predictions
βοΈ Regenerative cooling thermal analysis (Bartz correlation)
π― Multi-stage vehicle support (Falcon 9, Saturn V presets)
β‘ Validated against NASA CEA with <2% error β’ Numba JIT for 10-100x speedup
π§ͺ 226 automated tests with pytest β’ Published on PyPI
Advanced Real-Time Satellite Tracking & Orbital Analysis Platform
π 25,000+ satellites & debris real-time tracking with SGP4 propagation
π‘ Doppler Shift calculation for radio frequencies
π Orbital Decay prediction algorithms
β οΈ Conjunction analysis for collision warnings
π Pass Prediction algorithm for observer locations
π± PWA support & AR mode with compass-based guidance
High-Fidelity NEO (Near-Earth Object) Defense Simulator
π¬ N-Body gravity simulation with Velocity Verlet integrator (Rust)
π°οΈ NASA NeoWs API integration for real asteroid data
π Monte Carlo collision probability analysis
π₯ Kinetic Impactor deflection scenario modeling
π¬ Cinematic physics-based 3D rendering with Three.js
IEEE Std 686-2008 Compliant Radar Simulation & Operator Console
π‘ Swerling I-IV fluctuation models & ITU-R P.676 atmospheric attenuation
π― LFM/Barker waveforms, CA-CFAR thresholding, SAR/ISAR imaging
π€ Random Forest ML for target classification (Drone/Fighter/Missile)
π₯οΈ 30+ FPS PPI Scope, A-Scope & Range-Doppler displays
π Cross-platform builds (Windows, macOS, Linux) with GitHub Actions CI/CD
Full-Stack Turkish Lunar Mission Simulator
20+ REST API endpoints β’ Real-time 3D orbital mechanics β’ Integration testing
WebGL Missile Guidance & 6-DOF Simulation
RK4 physics integration β’ Proportional Navigation guidance β’ Real-time control loop
Interactive Geography Game Platform
Google Maps/Leaflet integration β’ XP/Level algorithms β’ Local caching optimization
| Principle | Description | |
|---|---|---|
| π― | Design First | Understand the problem before writing code |
| π¬ | Simulate | Physics-based models over black-box assumptions |
| π | Measure | Quantified uncertainty & error bounds |
| β‘ | Profile | Benchmark critical paths |
| π‘οΈ | Fail Safe | Graceful failure, never silent |
| Domain | Technologies |
|---|---|
| Physics | 6-DOF dynamics, RK4/Verlet integration |
| State Est. | Kalman Filter, EKF, UKF |
| Control | PID, Adaptive, Gain Scheduling |
| Analysis | Monte Carlo, Genetic Algorithms, Pareto |
| Guidance | Proportional Navigation, Missile dynamics |
| DSP | Monopulse tracking, Doppler, ECM |
Core Competencies:
- πΉ 6-DOF rigid body dynamics & kinematics
- πΉ Sensor fusion & calibration (SOLT method)
- πΉ Real-time control loop design
- πΉ Hardware-in-the-loop (HIL) testing mindset
- πΉ Monte Carlo uncertainty propagation
| Domain | Focus |
|---|---|
| LLM | RLHF, Evaluation pipelines, Hallucination analysis |
| ML | KNN, SVM, Decision Trees, Classification |
| Optimization | Genetic Algorithms, Multi-objective (Pareto) |
| Data | Feature engineering, Synthetic data generation |
| Area | Technologies |
|---|---|
| MCU | ESP32, Arduino, ARM Cortex-M, Raspberry Pi |
| Protocols | UART, I2C, SPI, CAN |
| Control | Adaptive PID, Gain Scheduling, Flight Controllers |
| Tools | Altium Designer, Logic Analyzer, RTOS concepts |
Highlights:
- πΉ Rover & robotic arm embedded architecture
- πΉ Motor driver control optimization (C/C++)
- πΉ PocketVNA signal processing
- πΉ FPV drone flight controller logic
| Certificate | Institution |
|---|---|
| π°οΈ Digitalisation in Aeronautics & Space | TUM (Technical University of Munich) |
| π‘ IoT Networking (Honors) | University of Illinois |
| π» IT Essentials: PC Hardware & Software | Cisco Networking Academy |
| Skill | Level |
|---|---|
| π§ AI/ML & LLM | β¬β¬β¬β¬β¬β¬β¬β¬β¬β¬ 95% |
| π Simulation & Physics | β¬β¬β¬β¬β¬β¬β¬β¬β¬β¬ 97% |
| π» Full-Stack Dev | β¬β¬β¬β¬β¬β¬β¬β¬β¬β¬ 90% |
| βοΈ Embedded Systems | β¬β¬β¬β¬β¬β¬β¬β¬β¬β¬ 85% |
| π‘ Radar & DSP | β¬β¬β¬β¬β¬β¬β¬β¬β¬β¬ 80% |
| π¨ 3D/WebGL | β¬β¬β¬β¬β¬β¬β¬β¬β¬β¬ 90% |
| π± Mobile (Flutter) | β¬β¬β¬β¬β¬β¬β¬β¬β¬β¬ 85% |
| Language | Level |
|---|---|
| πΉπ· TΓΌrkΓ§e | Native |
| π¬π§ English | Intermediate (B1-B2) |
"First, solve the problem. Then, write the code." - John Johnson
β Built with π by SpaceEngineerSS
