ai/ml engineer @ turkcell technology
utku = {
"role": "ai/ml engineer @ turkcell technology",
"studying": "m.sc. data science & ai @ boğaziçi university",
"graduated": "b.sc. ai engineering @ tobb etu",
"building": "llm-powered ai call center backend",
"stack": ["langgraph", "langchain", "litellm", "langfuse", "mcp",
"whisper", "faiss", "fastapi", "postgresql", "redis",
"opensearch", "openshift", "jenkins"],
"research": "deepfake detection in biometric systems @ ieee asyu 2025",
"fuel": ["sleep", "caffeine", "l-citrulline"],
}- AI-Powered Deepfake Detection in Biometric Systems built a domain-independent deepfake detection pipeline for biometric security systems. combined multiple manipulation sources into a hybrid dataset, reaching high accuracy and strong generalization on both known and unseen data.
ai_ml: [llms, langgraph, langchain, langfuse, litellm, google-adk, mcp, whisper, faiss,
computer-vision, yolo, deep-learning, tensorflow, pytorch, scikit-learn, opencv]
backend: [fastapi, flask, sqlalchemy, postgresql, redis, opensearch, mysql, sqlite]
devops: [openshift, kubernetes, docker, jenkins-ci/cd]
cloud: [gcp, vertex-ai, cloud-run]
edge_ai: [model-optimization, tflite, onnx, raspberry-pi, arduino]
languages: [python, c, c++, java, c#, r]






