• Architected an auto-grad engine from the ground up to build deep neural networks exceeding 200000+ parameters and 100+ layers, achieving backpropagation on par with top libraries.

Github: https://github.com/PrateekJannu/Gradient_Engine

• Engineered a real-time database application with Microsoft Power BI, consolidating 45 tables and 500,000+ rows of UDL data

• Constructed an independent end to end software using Python 3.6 currently used by 5+ industries automating the process of reading data files from actuators used by the industries, saving 3+ hours of field testing, software built was also field tested and accepted as the standard for reading data files increasing efficiency.(Increased efficiency by 62 %) 

GitHub : 

 https://github.com/PrateekJannu/AUMA/blob/main/aimsev_withoutspace2.p

• Installed another automation tool to convert BMP files to C code reducing 2 days of manual work to 20 minutes of program runtime, helping AUMA to develop products in regional languages such as Chinese.(Efficiency boosted by 41 %) 

https://drive.google.com/file/d/1HQXKoor1HzxhvdeJXn1WmL93X12OBQdU/vi

Undergraduate Research Assistance, Corporate Partnership Program 

• Increased program efficiency by creating and performing a tracking system for all CPP activities. 

• Constructed a website for the career fair with over 8,000 website visits. 

Web dev:  https://www.purdueece.com  (Development Percent 100%)

Secured a $9000 prize at a startup pitch competition with the startup professor. the hiring of students for research has been made easier by https://www.professor.wiki