
Competing at TEKNOFEST 2026
Serhan EnsarBüdün
Computer vision · Embedded systems
Ege University — Computer Engineering, 2024–2029
Scroll ↓
01About
Serhan is a Computer Engineering student at Ege University in İzmir, working on AI-driven robotics, computer vision and embedded systems.
Right now he captains two TEKNOFEST 2026 teams — LAÇİN and TUYGUN — and leads embedded work on two more. Four competition projects running at once.
As Computer Society Chairperson at IEEE Ege University he leads six technical sub-teams and teaches Arduino and embedded systems to other students.
He builds end to end: Deneyap firmware at one end, React and Electron ground stations packaged as standalone executables at the other.
- mAP@50, LAÇİN
- 0.655
- Labelled images
- 116K
- Concurrent TEKNOFEST teams
- 4
- Sub-teams led at IEEE
- 6
02Capabilities
Computer Vision
88Detection and tracking pipelines, from raw labels to a deployed model
- YOLOv8 / v11
- ByteTrack
- SAHI
- OpenCV
- Roboflow
- GMC
Python & AI
90Training pipelines, data tooling, desktop GUIs and API layers
- Python
- FastAPI
- Flask
- CustomTkinter
- PyInstaller
- Blender bpy
Embedded & IoT
85Firmware where timing, sensors and radio links have to hold together
- C++
- ESP32
- Deneyap
- Arduino
- Raspberry Pi
- STM32
- IMU
- RFID
- LoRa
Interfaces
78Ground stations and dashboards that stay readable under load
- React
- TypeScript
- Vite
- Electron
- Kotlin
- Jetpack Compose
- Nextion HMI
03Work
4 Active · 3 Complete
LAÇİN
ActiveAerial AI Object Detection System
mAP@50 = 0.655 across 116K images
2025–2026
Team Captain
TUYGUN
ActiveAerial AI System
Visual odometry for when GPS drops out
2025–2026
Team Captain · Systems Integration
EGE ODBARS
ActiveAutonomous Unmanned Ground Vehicle
6×4 rocker-bogie platform with a live ground station
2025–2026
Embedded Software & Computer Vision
EGENODE
ActiveDynamic Logistics Mobile Robot
RFID, IMU and a 4-DOF arm on one 240 MHz dual-core
2025–2026
Embedded Software & HW-SW Lead
HomeAgent
CompleteAI-Powered Smart Home Ecosystem
One Raspberry Pi hub, four client surfaces
2025
Solo Developer
TELEMETRY
CompleteSmart Telemetry System
Wear OS to ESP32 to Nextion HMI at 1 Hz
2025
Solo Developer
TÜBİTAK
CompleteElectric Motor Innovation
The first project — motor efficiency, age 14
2019
Co-Researcher
04Roles
03 · Sept 2024 — present
IEEE Ege University Student Branch
Computer Society Chairperson
- Leads the Computer Society committee and its six technical sub-teams
- Designed and taught a five-week Arduino and circuit design course in C++
- Coordinated branch events and represented the university at CSCAMP and the Ege Regional Meeting
02 · 2025 — 2026
TEKNOFEST 2026
Team Captain ×2 · Embedded Lead ×2
- LAÇİN and TUYGUN — aerial AI systems: airborne detection and GPS-denied visual odometry
- EGE ODBARS — autonomous UGV, the ODBARS NEXUS ground station, synthetic dataset generation
- EGENODE — logistics robot, Deneyap firmware, RFID with IMU and a 4-DOF arm
01 · 2024 — 2025
IEEEXtreme · Google DevFest
Competitor · Attendee
- IEEEXtreme — the global 24-hour algorithmic programming competition
- Google DevFest — sessions with industry engineers on emerging technology
Certifications
IEEEXtreme 19.02025
IEEE
Global rank 1453 · Team Egelcarus
Savunma Sanayii 1012026
SSB × YÖK
Defence-industry programme · grade BA
Occupational Health & Safety2026
EGESEM · Ege University
16-hour certified programme
Simulink Fundamentals2026
FİGES · MathWorks
MATLAB & Simulink training
05Internship
Boğaziçi Teknopark
Entrepreneurship Office · Summer Intern
June 2026 — present · İstanbul
Turned manual startup research into an autonomous pipeline
Serhan was handed a recurring task: research startups across the Turkish and US markets — filter announcements by date range, back-analyse them, trace each founder's prior experience, and report it all. The obvious answer was to grind through it by hand in a spreadsheet. He built the pipeline that does it without him instead.
The pipeline
- 01 Scrape sources
- 02 Filter & summarise
- 03 Fill the gaps
- 04 Structure to JSON
- 05 Categorise
- 06 Auto-fill documents
n8n orchestrates the flow; the ChatGPT (GPT-4o) API turns filtered text into clean, structured JSON. Before committing to it he benchmarked six language models on cost and rate limits and settled on the one that kept the run economical — by then it had already cleared the first 80 rows of a 440-row backlog on its own.
- n8n
- ChatGPT API
- GPT-4o
- Automation
- JSON
Are You Ready?
You can continue exploring the portfolio or embark on a cinematic journey.

Let's build something
Serhan is open to work on robotics, computer vision and embedded systems. Deep in TEKNOFEST 2026 right now, but always reading messages.



