ML Engineer – Threat Detection Models
Core
Build real-time ML models to detect security risks (prompt injection, jailbreaks, data exposure) in AI agents and LLM applications.
Role type
Senior ML Engineer (AI Security)
Builds
Low-latency threat detection models serving in Go-based gateway and endpoint services
Domain
Cybersecurity / AI Security / Enterprise SaaS
Deliverable
production ML models
Required skills
Python, Go, PyTorch, Hugging Face Transformers, scikit-learn, model fine-tuning, low-latency inference optimization, MLOps, Docker, Kubernetes, cloud platforms
Preferred skills
Prompt injection detection, DLP/PII classification, adversarial ML, AI-assisted development workflows
Technologies
ONNX Runtime, TorchServe, Triton, vLLM, Claude Code, Cursor, GitHub Copilot
Responsibilities
Design and train threat detectors using classifiers and fine-tuned LLMs; build and maintain training datasets and evaluation benchmarks; serve models under strict latency constraints; optimize inference via quantization and distillation; integrate with security researchers to update detection logic; manage MLOps pipelines for model rollouts.
Seniority
Senior, hands-on IC