At Vocab.ai, I developed and supported production AI services for ASR, speaker diarization, and model inference. I deployed AI workloads with Docker on Linux and GPU-based environments.
I supported open-source model deployments, including Whisper and NVIDIA NeMo, and prepared multilingual speech datasets for model development and evaluation. I also evaluated speech models using Word Error Rate (WER).
At Fisec Global Inc., I containerized Python, Java, and frontend applications with Docker. I developed and maintained Jenkins CI/CD pipelines for automated builds, testing, packaging, and deployment.
I also worked with Harbor and Gitea, and configured Nginx reverse proxies, SSL, and Linux application environments for production deployments.
