NVIDIA
Apr 2021 - Present
Senior AI Software Engineer (Apr 2024 - Present)
Warsaw, Poland
- Working with cross-organizational teams, including engineers, researchers, and marketing, to deliver NVIDIA Inference Microservices (NIMs) for domains like Retrievers, Computer Vision, and Large Language Models (LLMs).
- Optimizing and deploying state-of-the-art models from the community, internal research teams, and partners to build.nvidia.com, used by Fortune 500 companies worldwide.
- Developing a prototype tool to automate the performance optimization and deployment of unoptimized Huggingface checkpoints into NIMs.
AI Software Engineer (Mar 2022 - Mar 2024)
Warsaw, Poland
- Performance optimized existing GNN and Transformer-based Recommendation Systems implementations (up to 2000x faster than original research implementations).
- Developed computer vision models and exploring new approaches for NSFW content filtering.
- Optimized and deployed LLMs to AI Foundational Models catalog (now NVIDIA NIM).
AI Software Engineering Intern (Apr 2021 - Mar 2022)
Warsaw, Poland
- Researched and documented state-of-the-art Recommender Systems algorithms for the NVIDIA DeepLearningExamples repository.
- Optimized research models, achieving up to 108x speedups in workloads, and tuned performance on industrial GPUs like DGX-1, DGX-2, and DGX-A100.
- Contributed to the development of models such as SIM/TF2.