At Zoom Video Communications, I investigate production escalations across application, operating system, networking, client-server, and VDI layers. I’ve resolved approximately 2,800 issues for Premier and Premier Elite customers, using log analysis, network traces, and root-cause analysis.
I developed VISTA self-service troubleshooting and contributed to ATLAZ incident-analysis automation. VISTA contributed to 22% fewer support tickets, while ATLAZ improved issue resolution time by 28%.
I co-developed an agentic AI platform at Zoom using TypeScript and NestJS microservices. My contributions included Kubernetes-optimized deployment, MongoDB checkpoints for fault recovery, and SSE workflow monitoring.
Earlier, as a Software Engineering Intern at NVIDIA, I developed Hugging Face Transformer punctuation models and WFST text normalization for speech recognition. I also contributed to Russian Conformer training.

