Resume

TechCorp AI Labs

• •Developing novel attention mechanisms for vision transformers, achieving 15% accuracy improvement with 40% fewer parameters
• •Implementing efficient inference pipelines using TensorRT and ONNX for edge deployment
• •Collaborating with senior researchers on multimodal learning research for upcoming CVPR submission

Stanford AI Lab

• •Contributing to research on multimodal learning and efficient transformer architectures
• •Developed data processing pipeline handling 10TB+ of image-text pairs for pretraining
• •Co-authored 2 papers on efficient attention mechanisms accepted to top-tier venues

StartupAI

• •Built real-time object detection system achieving 60fps on edge devices using TensorRT optimization
• •Developed MLOps pipeline with automated training, evaluation, and deployment using Kubernetes
• •Reduced model inference latency by 65% through quantization and kernel fusion techniques

Hugging Face

• •Contributing to the transformers library with focus on model optimization and inference
• •Implemented support for new model architectures including LLaMA and Mistral variants
• •Authored documentation and tutorials with 50K+ views on efficient fine-tuning techniques

Stanford University

• •TA for CS229 (Machine Learning) and CS231N (Computer Vision)
• •Conducted weekly office hours helping 200+ students with assignments and projects
• •Developed new programming assignments on attention mechanisms and transformers