Delivering a sustainable application to securely manage sensitive device settings using certificates & public key cryptography, preventing security exposure and other password related pitfalls

Constructed and optimized own training pipeline including data augmentation in PyTorch while working with the security team to reduce training loop by 50% and achieve classification accuracy of over 90%

Engineered a new backward compatible framework for collated copies using Linux development tools (diff, vim, gitk) and OOPS concepts that advertises parameters to existing firmware from EWS which enabled new features

Worked on a model using Python for ‘Identifying the Relative Importance of Customer Issues on Product Ratings’ and coupled Long Short-Term Memory (LSTM) units of a Recurring Neural Network with Attention Model with Machine Learning team

Created a NVM backed interface to select monochrome print quality, facilitating value to be detected dynamically enhancing customer experience, code quality and performance

Learned and adopted new technologies and tools of firmware to program 1000+ modifications to 4 central HP Firmware repositories in less than 3 months working in a team environment collaboratively