Kaan Donbekci
Machine Learning Research Assistant at Columbia Spoken Language Processing Group
San Francisco, California, United States
Summary
👤
Senior🎓
Top SchoolKaan Donbekci is a machine learning engineer and current MSCS student at Columbia University with eight years of professional experience spanning research and product-focused roles. He develops cross-lingual, multimodal emotion-detection models using transformer-based transfer learning and self-supervised speech representations (Wav2Vec2, RoBERTa, FAb-Net) on a DARPA-funded collaboration across three labs. Previously at Neosensory he built low-latency, privacy-preserving speech recognition for an embedded wristband and led ML infrastructure and two interns, with earlier work compressing and deploying models to microcontrollers. His background in symbolic systems from Stanford and hands-on IoT and behavioral study tooling gives him a rare blend of human-centered research and production ML engineering. He’s currently exploring audio-first data management and MLOps practices to streamline model building across multimodal pipelines.
8 years of coding experience
4 years of employment as a software developer
Master of Science - MS, Computer Science, Master of Science - MS, Computer Science at Columbia UniversityHigh School, High School at Robert CollegeBachelor of Science - BS, Symbolic Systems, Bachelor of Science - BS, Symbolic Systems at Stanford University
Turkish, English, French
Stackoverflow
81reputation
3kreached
0answers
3questions
Github Skills (56)
python10
data-science10
machine-learning10
ml10
deep-learning10
tensorflow10
deep-neural-networks10
neural-network10
speech9
data-augmentation8
virtual-screening8
datasets8
webdataset8
jupyterlab8
evaluation7
Programming languages (6)
Github contributions (5)
kdonbekci/kd
Jul 2017 - Mar 2021
Repository for my personal website in works.
Contributions:4 PRs, 70 pushes, 7 branches in 3 years 9 months
PervasiveWellbeingTech/Subliminal-Room
Jul 2018 - Oct 2018
Contributions:109 pushes, 1 branch in 3 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.