At Cisco, I build software automation tools that centralize defect tracking and build monitoring. I developed a team dashboard integrating CDETS and Jenkins, with filters and deep links to source pages.
I fine-tune large language and video-language models, including DeepSeek-R1-Distill-Qwen-14B and Qwen2.5-VL, using QLoRA on NVIDIA A40 GPUs. I build Python training and inference pipelines with PyTorch, Hugging Face Transformers, video preprocessing, and dataset preparation.
I also evaluate GPU utilization, latency, and inference throughput, and have ported Intel GPU inference scripts to NVIDIA A40 environments. I'm a Computer Science graduate focused on software automation, CI/CD integration, and practical AI model deployment.
