Hello, I'm

Öznur Karahasan

Full Stack Developer & IoT Engineer

About Me

I am a developer who loves bridging the gap between software and the physical world. I specialize in building AI-powered assistants, mobile applications with Flutter, and embedded systems using Raspberry Pi/Arduino.

From creating Edge AI presentation assistants to fine-tuning vision-language models and building local agent ecosystems, I focus on solving complex problems with technologies like Python, MS-Swift, and LangChain.

Tech Stack

GitHub Contributions

My digital footprints in open source

Activity Graph

Less

More

Connected as:oznurkarahasan

* Reflects contribution activity from the past year.

Featured Projects

Live Translate (Rust & AI)

An ultra-low latency, real-time speech transcription and translation system. Built with Rust for high-performance audio processing, it features real-time subtitle delivery and WebSocket-based AI integration.

AI-Powered Presentation Assistant

An intelligent system that analyzes real-time speech during presentations to automatically navigate slides, powered by Edge AI logic and a modern web stack.

Local Agent Collective

A scalable ecosystem of multiple AI agents designed to run fully locally. It ensures total data privacy by utilizing local LLMs through Ollama and LangChain for task orchestration, eliminating the need for external cloud dependencies.

Analog Clock with ESP32 & NeoPixels

ESP32-based analog clock with NeoPixel LEDs. Features NTP time sync over WiFi, RTC backup, web interface control, LDR brightness auto-adjustment, and OTA updates.

Flight Tracking

Automated flight price monitor that tracks airfares and sends email alerts upon price drops. Features Docker containerization and GitHub Actions for scheduled execution.

Danilo: Voice Kitchen Assistant

An AI-powered voice assistant for the kitchen. Uses Adaptive RAG architecture and FAISS for semantic search to provide personalized recipe guidance.

LLM fine tuning with MS-Swift

End-to-end fine-tuning and deployment of the Qwen2-VL-7B-Instruct vision-language model using MS-Swift for high-accuracy traffic sign recognition.

Wastewater Predictive Maintenance

Machine Learning project focused on predicting maintenance needs for wastewater treatment. Analyzes sensor data to prevent system failures.

Vehicle Tracking & Mapping (C++)

Real-time tracking and mapping system. Delivers accurate location data, route planning (A*), and drone integration support.

Indoor Navigation (AR/BLE)

Flutter-based indoor navigation system for visually impaired. Uses BLE signals for detection and AR technology for guidance.

DualFlow

A virtual fluid physics engine simulating a 'digital water balance'. Responds to inertia and momentum through motion-based input.

Restaurant Automation (Laravel)

Full-stack management system with role-based access, real-time stock tracking, and secure authentication.

CineMood

Web application for discovering and recommending movies via TMDB integration. Features personal list management and genre filtering.

Student Grade System

WinForms application for managing educational data using Entity Framework 6 (Database-First).

My AI Portfolio

This interactive Terminal/Shell based portfolio built with modern web technologies.

Latest Articles

Get in Touch

Have a project idea involving AI or IoT? Let's collaborate.

Follow Me: