Haleema Sadia
Deep Learning Software Engineer at Intel Corporation
Sunnyvale, California, United States
Summary
🤩
Rockstar🎓
Top SchoolHaleema Sadia is a Deep Learning Software Engineer with 7 years of experience building production-grade ML and systems software at Intel, combining strong research instincts from a Purdue MS with hands-on open-source contributions. She has extended core primitives in the oneDNN library—adding new activation functions, bfloat16 support, and 2D/3D deconvolution—demonstrating a focus on performance-critical kernel work that impacts ML runtimes. Her background spans end-to-end projects from smartwatch sensor attacks to real-time mobile accessibility apps, showing an ability to move ideas from prototype to deployable systems. A collaborative teacher and mentor who has led university courses and Girls Who Code classes, she pairs technical depth with clear communication. Outside engineering she pursues photography and philosophy, bringing a curious, outdoorsy perspective to problem solving.
7 years of coding experience
2 years of employment as a software developer
Bachelor of Science (B.S.), Software Engineering, 3.82/4.00, Bachelor of Science (B.S.), Software Engineering, 3.82/4.00 at National University of Science and TechnologyMaster of Science - MS, Computer Science, Master of Science - MS, Computer Science at Purdue University
Github Skills (8)
neural-network10
oneapi10
c-language10
deep-learning10
edn10
cprogramming-language10
deep-neural-networks10
opencl9
Programming languages (1)
Github contributions (1)
uxlfoundation/oneDNN
Apr 2019 - Nov 2022
oneAPI Deep Neural Network Library (oneDNN)
Role in this project:
Back-end Developer
Back-end DeveloperContributions:23 reviews, 159 commits, 9 PRs in 3 years 7 months
Contributions summary:Haleema extended the eltwise primitive within the oneAPI Deep Neural Network Library (oneDNN), adding support for several new activation functions and bfloat16 data type. These changes involved modifying the core kernel code to implement the forward and backward passes for the new operations and included test updates. The user also implemented 2D and 3D deconvolution in floating point 32, further enhancing the library's functionality. The contributions focused on performance improvements.
bfloat16sse41avx512openmpopencl
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.