Sachin Agarwal

Ml Infrastructure & Security Engineering Lead Roboflow (y C Series-b) @Roboflow

Waterloo, ON, CA

Architected and scaled serverless inference on Kubernetes/Knative (distributed file system r/w cache, dynamic GPU memory management), supporting millions of daily requests. Designed and deployed infrastructure-as-code (Terraform, Helm) for node GPU clusters for AI model training/inference (YOLO, CLIP, VLMs). Led 3-person SRE team, transitioning to proactive, KPI-driven reliability. Launched Video Inference and Automated Image Labeling (zero/few-shot learning) features. Achieved SOC 2 Type 1 & 2 certifications (3 annual runs), implementing robust security controls Reduced annual cloud costs >$1M via optimization (spot instances, storage tiers). Implemented CI/CD (GitHub Actions), enhancing developer productivity. Managed Elasticsearch, Firestore, and GCP/AWS infrastructure. Developed automated ML-Ops framework for model lifecycle management. Automated vulnerability remediation (SAST, GitHub Dependabot).

Boston University

PhD, Computer Engineering

Boston University

M.S., Computer Engineering

National Institute of Technology Warangal

Bachelor of Technology, Electronics and Communication Engineering