Nov 2021 – Jan 2025 · 3 yrs 2 mos

• Key Responsibilities:
• Directing the development and implementation of robust healthcare AI solutions.
• Managing and mentoring a team of skilled engineers.
• Overseeing Digital Twin platform projects in partnership with LX.
• Creating scalable technological solutions for Kia Motors and LG Electronics.
• Conducting research and development in deep learning, computer vision, and biosignal processing.
• Key Achievements:
• Published 28 research papers in top-tier journals and conferences.
• Developed AI models for medical imaging and commercial applications.

Mar 2021 – Oct 2021 · 7 mos

• Advising the Company with respect to the End-to-End Development of Software Applications.

May 2019 – Feb 2021 · 1 yr 9 mos

• The Terenz simplifies the diagnosis process led by Physicians using Deep Learning Techniques by providing a weighted decision based on medical images.
• Supporting the company on the "Voluntary" basis for the development of the Deep Learning Models.
• Presently, participating in the research regarding the analysis of Brain MRI and CT Scans for the prediction and forecasting of Neurodegenerative Diseases such as Alzheimer's and Parkinson's.
• Projects:
• Developed Cov-XR AI for the analysis of Chest X-Rays for the detection of COVID.
• Developed Deep Learning Models for the product "Mellius" which allows the prediction and grading of Diabetic Retinopathy and also localization of aneurysms in Retinal Fundus Images.
• Developed Reseption (3D ResNet 50 + 3D Inception V3) for the product "Demenor" which allows the prediction and grading of Alzheimer's Disease and also localization of aneurysms in MRI Scans using 3D Gradient Class Activation Maps.
• Developed Reseption (3D ResNet 50 + 3D Inception V3) for the product "Kinon" which allows the prediction and grading of Parkinson's Disease and also localization of aneurysms in MRI Scans using 3D Gradient Class Activation Maps.

Feb 2021 – Oct 2021 · 8 mos · Busan, South Korea

• PiXative - Vessels and Ship Detection Platform from Optical Satellite Imagery
• Development of Preprocessing Pipelines for Optical Geo Raster Scans [Satellite Imagery]
• Development of Tile Based Object Detection Algorithm using MaskRCNN with Reseption Backbone.
• Land Cover Detection Algorithm for discriminating between Land, Port, and Ocean.
• Development of Restful API Service for deploying the Model.

Sep 2018 – Sep 2020 · 2 yrs · Republic of Korea

• Biosignal Analysis using Machine Learning and Deep Learning
• Sleep Pattern Analysis using Accelerometer, Gyroscope, and Photoplethysmogram (PPG) Data. [Wavelet]
• State of the Mind Prediction using Accelerometer, PPG, ECG, EDA and Oximeter Data using MultiChannel CNN
• Activity and Emotion Detection for Household Pets [http://www.sweetsolution.co.kr]
• Pneumonia Detection from Lung CT Scans.
• Interpretability of Deep Learning Models using 3D-Gradient Class Activation Maps.
• Development of Systems for the Computer-Aided Detection of Neuro Degenerative Diseases (Alzheimer's and Parkinson) by using MRI Imaging
• Early Detection of Alzheimer’s Disease from 1.5 T MRI Scans Using 3D Convolutional Neural Network
• 3D Textural, Morphological and Statistical Analysis of Voxel of Interests in 3T MRI Scans for the Detection of Parkinson's Disease Using Artificial Neural Networks
• Detection of Parkinson's Disease from 3T T1 Weighted MRI Scans Using 3D Convolutional Neural Network

Apr 2018 – Aug 2018 · 4 mos

• HighRadius is a Fintech enterprise Software-as-a-Service (SaaS) company. The HighRadiusTM Integrated Receivables platform optimizes cash flow through automation of receivables and payments processes. Powered by the Rivana Artificial Intelligence Engine and Freda Virtual Assistant for Credit-to-Cash.
• Worked specifically on the Collections Artificial Intelligence Platform of the company and deployed Machine Learning Models for the Prediction of the Payment Date of an Invoice and Promised to Pay of a an invoice.

May 2017 – Apr 2018 · 11 mos

• Working for the Artificial Intelligence platform ‘Rivana’ for supporting the proprietary software Deduction Application and Collections Application of the company. Developed a Generic Machine Learning Model to predict valid or invalid deductions of an invoice. Also worked upon the Collections Platform to Predict the Payment Date or the Clearing Date of an invoice. Developed an algorithm to convert the Categorical Data or Features to a Numerical Form using the Average Response Ratio from the Training Set corresponding to the true labels.