Experience

3,274 controlled trials on indirect prompt injection across LangChain, MCP, and the Anthropic API. Preprint submitted. Dataset open-sourced on GitHub and Zenodo.

IAM at scale. Terraform across 80K+ GCP role bindings, FedRAMP compliance, and an AI-assisted workflow tool. Top-performing intern of 120+.

Cloud security consulting for healthcare clients. AWS migration assessment, IAM analysis via NIST CSF, and SAST on internal apps.

Red teaming focused on internal penetration testing of aerospace systems. Presented findings directly to CISO, resulting in SDLC changes to secure flight-critical software.

Appsec and ML. Found path traversal and IDOR in banking apps. Built a fraud detection model on 2M+ transactions and an ML spam filter at 93% accuracy.