Jyotika Yadav

Computer Vision Developer

Machine Learning Engineer



Data Scientist at ConcertAI (Terarecon) - Current
  • Led the creation of a segmentation-based localization model targeting critical Aorta segments within CT volumes, integral to the TAVR/EVAR solution package, awaiting FDA approval.
  • Engineered a multiorgan segmentation model achieving a remarkable 92% Dice score, accurately labeling 27 distinct organs in full-body 3D CT scans, deployed in the new Life Sciences platform.
  • Developed a proof of concept for identifying CT scans with Pulmonary Embolism, leveraging classification techniques and Grad-CAM.
  • Currently developing a novel approach for data selection optimization, aimed at reducing dataset requirements while maintaining peak performance standards for segmentation models, utilizing autoencoders and clustering methods to expedite data cleaning and annotation processes.
AI Engineer at Sigtuple Pvt Ltd
  • Built an end-to-end ML pipeline for a blood smear analyzer, reducing analysis time by 85%, from 2 minutes to 20 seconds.
  • Coordinated the FDA approval process of the blood smear analyzer, conducting data analysis, system design documentation, and priority bug fixing.
  • Implemented a WBC detector for Microfluidics blood chambers, providing total counts and differential count with 95% accuracy.
  • Collaborated with IISc. Bengaluru Professor, Dr. Prathosh A. P., and pathologists to research the identification of Malarial Parasites in Peripheral Blood Smear.
Freelancer for Unify Systems
  • Collaborated with a client to develop a segmentation model aimed at identifying the incision point for tracheostomy from ultrasound images.
  • Managed the project from inception to deployment, overseeing data collection and model development.
  • Successfully delivered the project on schedule and surpassed client expectations.