LLM Applications & Prompt Security Engineer
Join the engineering team building hardened digital infrastructure, zero-trust cryptographic platforms, and high-velocity developer tools.
Role Overview
Guard generative AI deployments against prompt injection, jailbreaking, PII leakage, and non-deterministic hallucination. You will engineer runtime LLM guardrails, automated red-teaming benchmarks, and structured output verification layers.
Core Responsibilities
- → Implement runtime defensive guardrails (NeMo Guardrails, Llama Guard) to block prompt injections and adversarial inputs.
- → Build automated red-teaming pipelines that stress-test LLM agents against unauthorized tool executions and system prompt exfiltration.
- → Engineer deterministic JSON schema enforcement using instructor, outlines, or native structured output APIs.
- → Design synthetic benchmark suites to evaluate retrieval relevance, faithfulness, and semantic drift across model updates.
- → Monitor production token consumption, latency budgets, and model safety telemetry.
Key Requirements & Qualifications
- ✓ 1 to 3 years of experience in software engineering with a focus on LLM application security or NLP pipelines.
- ✓ Familiarity with OWASP Top 10 for LLMs (prompt injection, sensitive information disclosure, insecure output handling).
- ✓ Hands-on experience with Python, Pydantic, and structured output generation libraries.
- ✓ Working knowledge of embedding similarity metrics and automated factuality evaluation frameworks.
- ✓ Strong curiosity for emerging adversarial AI security research and red-teaming methodologies.
Why Build With EyesOn SafeTech
No micromanagement or endless status meetings. High trust with direct production ownership.
Solve difficult real-world engineering problems across cryptography, PKCS#11, and microservices.
Asynchronous communication, flexible work hours, and autonomy engineered from day one.
Competitive salary benchmarks, hardware stipends, and performance incentives.
Submit Your Application
Upload your resume to auto-fill details, or fill the form manually.
We extracted your details below. You can review and edit before submitting.
Application Received
Thank you for applying for LLM Applications & Prompt Security Engineer. Your application and resume have been delivered directly to our hiring team at hr@eyesonsafetech.com. We review all applications within 48 hours.