Cuda Based CNN Speed Optimization

I'm looking for a CNN optimization expert who can enhance the speed of my current Cuda based CNN.

Key Responsibilities:
- Speed up current Cuda based CNN optimization
- Integrate advanced architectural changes to enhance speed
- Incorporate EfficientNet into the existing architecture

Ideal Skills:
- Extensive experience with Nvidia Cuda and CNNs
- Deep understanding of EfficientNet
- Proven track record of enhancing speed in CNNs
- Ability to integrate advanced architecture changes

Please provide examples of previous work that align with this description.
Paper writing - Enhance base paper with minimum of 7-8pages in IEEE paper format

Reference base paper - Enhanced CNN optimization by using task parallelism