← Back to About

Software Engineer

Oracle

Jun 2025 - Present • Austin, Texas • Current

Building production AI infrastructure for computer vision applications, delivering real-time inference at scale

Key Accomplishments

  • ▸ Computer vision models needed production deployment → Built Kubernetes-based inference pipelines → Serving 10k+ requests/day with <100ms latency
  • ▸ Manual ML deployment took 2+ days → Implemented CI/CD for model updates → Reduced deployment time to 30 minutes with automated testing
  • ▸ License plate detection required 99%+ accuracy → Optimized OCR pipeline with preprocessing → Achieved 99.3% accuracy in production traffic

Skills & Technologies

  • Python
  • PyTorch
  • Computer Vision
  • Kubernetes
  • OCI

What I do

Build production-grade AI infrastructure for computer vision applications, from model development to deployment at scale with real-time inference.

Why it matters

  • Enables real-time detection at scale (10k+ requests/day).
  • Reduces deployment friction for ML teams.
  • Delivers high-accuracy vision systems for production use.

How it works

  • Inference pipelines: Kubernetes-based serving for real-time processing.
  • CI/CD for ML: Automated testing and deployment for model updates (30min vs. 2+ days).
  • Computer vision: License plate OCR (99.3% accuracy), facial recognition, video stream processing.
  • Monitoring: Prometheus, Grafana for model performance and uptime.

Tech

  • ML: PyTorch, TensorFlow, OpenCV
  • Infrastructure: Kubernetes, Docker, OCI
  • Languages: Python, Go

My role

  • Designed scalable inference architecture.
  • Built CI/CD pipelines for ML models.
  • Optimized OCR systems for production accuracy.