Niloofar Zarif
Software Engineer at Motorola Solutions
Vancouver, British Columbia, Canada
Summary
🤩
Rockstar🎓
Top SchoolNiloofar Zarif is a software engineer specializing in AI infrastructure and performance optimization with nine years of combined industry and academic experience. She has delivered measurable speedups—72% inference velocity gains in an autonomy SDK and 3.7x acceleration for embedding-heavy models—by bridging systems engineering, ML, and low-level C/C++ work. Her contributions to NVIDIA-Merlin’s Transformers4Rec and multi-GPU PyTorch tooling demonstrate hands-on open-source impact on real-world recommendation and training pipelines. She builds reliable CI/CD and cloud-native delivery (Jenkins, GitHub Actions, AWS) to productionize models across cloud and embedded targets. Based in Vancouver with an MS in Computer Engineering from UBC, she brings a rare mix of profiling-driven bottleneck analysis and practical hardware-aware optimizations. Beyond production systems, she also taught quantum computing curricula and developed educator-focused tools, reflecting a passion for making complex tech accessible.
9 years of coding experience
5 years of employment as a software developer
Diploma Mathematics, Diploma Mathematics at National Organization for Development of Exceptional Talents (Sampad)Master of Science - MS Computer Engineering, Master of Science - MS Computer Engineering at The University of British ColumbiaBachelor's degree Computer Software Engineering, Bachelor's degree Computer Software Engineering at Sharif University of Technology
English, French, Turkish, Persian
Github Skills (11)
pytorch10
transformer10
nlp10
python10
recommender-system10
bert9
cuda8
datatables8
tabular8
datatable8
huggingface7
Programming languages (7)
Github contributions (5)
NVIDIA-Merlin/Transformers4Rec
Sep 2022 - Dec 2022
Transformers4Rec is a flexible and efficient library for sequential and session-based recommendation and works with PyTorch.
Role in this project:
ML Engineer
ML EngineerContributions:16 reviews, 27 commits, 10 PRs in 3 months
Contributions summary:Niloofar's contributions primarily focused on improving the `transformers4rec` library, specifically addressing issues related to model training, and distributed data processing. They fixed bugs related to multi-GPU training using `DistributedDataParallel`, implemented masking for next item prediction tasks, and refactored code related to flags and model outputs. The user also collaborated with other developers on various enhancements and bug fixes within the library.
xlnetbertlanguage-modelseq2seqrecommendation
UBC-ECE-Sasha/PIM-Embedding-Lookup
Apr 2020 - Aug 2022
Contributions:18 commits, 6 PRs, 85 pushes in 2 years 4 months
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.