Mounika Gude
About Mounika Gude
Mounika Gude is a Senior Research Engineer at Blaize, with a background in data science and engineering. She has experience working at Cognizant and Honeywell Automation, and holds patents for innovations in computer vision networks.
Current Role at Blaize
Mounika Gude serves as a Senior Research Engineer at Blaize, a position she has held since 2020. In this role, she focuses on developing innovative solutions in the field of computer vision and machine learning. Her work involves implementing advanced techniques to enhance the performance and efficiency of models used in various applications.
Previous Experience at Cognizant
Mounika Gude worked at Cognizant in two capacities. She began as a Programming Analyst from 2014 to 2017, where she contributed to various software development projects. Following this role, she served as an Associate from 2017 to 2019 in the Hyderabad Area, India, further developing her skills in programming and data analysis.
Experience at Honeywell Automation India Ltd
Mounika Gude was employed at Honeywell Automation India Ltd as a Data Scientist for a period of six months in 2019. During her time in Pune Hadapsar, she focused on data-driven projects that aimed to improve automation processes and enhance operational efficiency.
Educational Background
Mounika Gude completed her Bachelor of Engineering (B.E.) in Mechanical Engineering from MVSR Engineering College from 2010 to 2014. Prior to this, she studied at the Board of Intermediate Education, focusing on Maths, Physics, and Chemistry from 2008 to 2010. Her foundational education was completed at Jawahar Navodaya Vidyalaya, where she achieved the Central Board of Secondary Education certification in 2008.
Innovations and Patents
Mounika Gude has received patents for her innovations in auto and manual mixed precision techniques specifically designed for computer vision networks. She developed a tool that converts floating-point models to fixed-point models, significantly enhancing inference speed while maintaining minimal accuracy loss. Additionally, she has implemented advanced quantization techniques, including per-channel quantization and auto mixed precision, to optimize model performance on hardware.