Yukti Adya

Integrated MSSQL with PyODBC to enable full CRUD operations, implemented process validation and UI enhancements, and built a Grafana dashboard for real-time production line monitoring improving operational visibility.

Researched and implemented deep-learning methods for 3D pose estimation and activity recognition on omnidirectional top-view images, achieving a 3.67% improvement in 3D action accuracy.

Implemented ANN-based handwritten character classification in MATLAB achieving 92% accuracy and developed a deep learning pipeline with MMPose to improve top-view pose estimation accuracy by 3.6%.

Collaborated in agile teams to improve RBC Windows and web application satisfaction by 15% and reduced application issues by 20% through development, validation, and client support.

Built web-based admin applications for user/role management and dashboards, increasing user efficiency by 20% and improving catalog data maintenance by 15%.

Optimized data processing software to improve system efficiency by 20%, migrated VB6 applications to VB.net reducing errors by 30%, and boosted team productivity via training and documentation.

Master of Science in Information and Communication Systems with coursework and projects in computer vision, image processing, and machine learning; thesis on 3D pose estimation and activity recognition.

Grade: 2.0

Bachelor of Technology in Electronics and Communication Engineering covering applied mathematics, programming, digital image processing, and networking.