Problem-solving driven data scientist with a demonstrated history of developing machine learning-based solutions for achieving business goals using industrial datasets.
Developed a machine learning-based approach for the automation and improvement of the die inking process, currently performed through visual inspection and manual die selection
Formed the basis of a layout pattern learning framework for systematic defect identification and subsequent automatic test generation. This work contributed in securing a Semiconductor Research Corporation (SRC) grant with the Trusted and RELiable Architectures (TRELA) lab
Examined the IC defect database through statistical analysis & layout feature extraction and developed a layout template matching tool using existing infrastructure, which automated a previously time-consuming process
Integrated the trimming cost-reduction methodology in the production flow, demonstrating its practicality & scalability
Researched and developed a novel machine learning-based methodology for adaptive IC trimming, with experimental results showing a 50% reduction in trimming time