Image Quality Enhancer
4× super-resolution and deblurring for blurry, low-resolution photos: an RRDBNet generator fine-tuned on a custom motion/defocus-blur degradation pipeline.
Machine Learning Engineer (PhD) with 6 years of experience in research and industry, designing and deploying scalable deep learning systems in production.
My work powers document intelligence pipelines processing millions of documents annually, serving enterprise clients across regulated industries.
From multi-GPU training on A100/H100 infrastructure to Kubernetes-orchestrated inference at scale, I turn advanced research into robust, high-performance AI systems with measurable business impact.
4× super-resolution and deblurring for blurry, low-resolution photos: an RRDBNet generator fine-tuned on a custom motion/defocus-blur degradation pipeline.
Unsupervised anomaly detection as a production-style system: autoencoders learn normal patterns across ECG and network-traffic domains, with a REST scoring API, CI, and Docker.
Manufacturing QA tool that aligns 3D scans to CAD references with a from-scratch ICP, renders per-point deviation heatmaps, and tracks process drift with control charts.
Two-stage deep-learning pipeline for single-lead long-term ECG that cuts false-positive arrhythmia alerts while holding ~99% sensitivity.
Peer-reviewed research publications — see the full list on Google Scholar.
Got a cool idea, a weird problem, or a project that needs some ML magic? Let’s build it together!