AI Red Team Engineer: Break & Harden LLMs for Demos

White Circle; New York (NY); On-site; USD 130,000 - 210,000; Full time

Benefits offered by this job

Equity

Flexible time off

Language lessons

Learning & development

All hardware and tools

Team off-sites

Job summary

White Circle is an AI Safety company focused on safety, reliability, and optimization for AI systems. We test policies, run adversarial experiments, and build scalable evaluation pipelines.

We’re seeking an AI Red Team Engineer to own end-to-end adversarial testing of LLM-powered systems, automate attacks, and turn findings into clear reports for demos, security reviews, and sales conversations. You will script attacks, maintain an attack library, and drive product improvements.

Qualifications

• Background in QA automation, AppSec, API/security/pen testing, or bug bounty.
• Strong Python scripting skills and scripting for automation.
• Experience testing APIs, web apps, backends, or SaaS products.
• Hands-on with LLMs, prompts, system instructions, RAG, agents, and tool/function calling.

Responsibilities

• Red-team LLM-powered systems: chatbots, copilots, RAG pipelines and AI products.
• Test for jailbreaks, prompt injection, data leakage, unsafe outputs, and policy bypass.
• Write Python scripts to automate attacks, collect responses, and generate reports.
• Build and maintain an internal attack library with demos and test cases.
• Turn model failures into clear reports: reproduce, assess impact, and fix recommendations.
• Convert attacks into regression tests and product requirements.
• Track new red-team techniques and fold useful ones into tests.
• Produce strong, credible evidence for customer demos and security reviews.

Skills

Python scripting

QA automation

API security testing

LLM adversarial testing

Penetration testing

RAG pipelines

Burp Suite

Postman

Playwright

pytest

Tools

LangChain

LangGraph

LlamaIndex

Job description

White Circle is an AI Safety company focused on safety, reliability, and optimization for AI systems. We test policies, run adversarial experiments, and build scalable evaluation pipelines.

We’re seeking an AI Red Team Engineer to own end-to-end adversarial testing of LLM-powered systems, automate attacks, and turn findings into clear reports for demos, security reviews, and sales conversations. You will script attacks, maintain an attack library, and drive product improvements.