Batuhan Edgüer¶
AI Researcher & ML Engineer
Building end-to-end ML systems — from meta-learning research and LLM fine-tuning to production anomaly detection. M.Sc. researcher at Yeditepe University, focusing on meta-learning and high-performance ML systems.
What I'm Working On¶
- Meta-Learning Thesis — Multi-LLM annotation pipeline with QLoRA fine-tuning on Qwen3-8B, calibrated active learning across 500+ papers
- Depth Estimation Paper — Compact monocular depth architectures with synthetic stereo generation, benchmarked on KITTI & NYU Depth V2
- Teaching — 3+ years as TA in Data Structures, Software Engineering & AI lab sessions at Yeditepe University, mentoring 100+ students per semester
Featured Projects¶
-
Meta-Learning Research System In Progress
M.Sc. thesis — automated pipeline extracting structured metadata from AI/ML papers and recommending optimal model configurations via semantic search.
PythonFastAPIQLoRAQdrantCelery -
Depth Estimation Paper
Compact monocular depth estimation with synthetic stereo pair generation. Cross-dataset benchmarks across multiple architectures and loss functions.
PyTorchOpenCVtimmHuggingFace -
Bitcoin Price Prediction Completed
NLP-driven prediction combining sentiment analysis on news headlines with financial indicators. Benchmarked 18 ML models + CNN/LSTM architectures.
TensorFlowscikit-learnXGBoostNLTK
Looking for an ML engineer or research collaborator?
I'm open to collaborations in meta-learning, NLP, and high-performance ML systems.