Selected work across local-first AI, RAG systems, internal assistants, and full-stack products. The goal is not just to show repositories, but to make the engineering direction clear.
Project Direction
These projects focus on useful workflows: private LLM memory, document-aware assistants, retrieval pipelines, and full-stack interfaces that can be maintained after the demo.
Private assistants, memory, and local model workflows.
Retrieval, evaluation, and document-grounded answers.
Python, Django, FastAPI, and service integration.
Clear interfaces for repeated operational use.
Privacy-first memory infrastructure for local LLM assistants with persistent conversation context and configurable knowledge sources.
A community DeepSeek Harness plugin for read-only Hugging Face Hub model discovery.
A local-first, explainable AI agent framework. Features self-healing, detailed error diagnostics, interactive tool-calling traces, and custom dataset evaluation for Ollama & LM Studio.
A local LLM-powered internal assistant for support requests, documents, weather, and organizational knowledge workflows.
The official CLI and Python client for the Hugging Face Hub.
Python and C++ tool for detecting exact and near-duplicate images in the Airbnb duplicate image dataset using native file hashing and perceptual hash distance.
This is a repository of agent notebooks that contains studies on how to use various agent libraries with different tools and methods.
Mini SIEM platform for log ingestion, threat detection, alerting, and security event monitoring
Invoice AI Extractor is an AI-powered tool that extracts structured data from invoices automatically. It combines OCR and modern language models to transform unstructured invoice documents into clean, machine-readable formats such as JSON and CSV.
Production-oriented RAG for JavaScript and React apps with hybrid search, caching, conversation management, and evaluation.
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
CP-Turkish-Sentiment-analysis is a curated collection of datasets, data preprocessing tools, model implementations, and Jupyter notebooks for training, evaluating, and deploying sentiment analysis models on Turkish text
A project showcasing modern development practices and practical software delivery.
A project showcasing modern development practices and practical software delivery.
I'll generally share what I learn and code related to artificial intelligence here.
A project showcasing modern development practices and practical software delivery.
This Streamlit application pulls current machine learning articles from arXiv and displays them to the user. Users can translate the titles and abstracts of articles into Turkish, like the articles, and view GitHub links.
I’m open to roles, collaborations, and freelance projects around LLM applications, local-first AI, RAG systems, and intelligent web products.