AI Red Teamer
Core
Develop and run adversarial test suites (manual and scripted) for LLMs and image/video models, crafting multilingual prompts and jailbreaks to identify policy edge cases.
Role type
AI Red Teamer
Builds
Adversarial test suites, vulnerability reports, and internal tooling (prompt libraries, scenario generators, dashboards)
Domain
AI Safety / Adversarial Machine Learning
Deliverable
production ML models
Required skills
Adversarial testing, prompt engineering, vulnerability analysis, scripting (Python/Bash), multilingual communication, technical writing
Preferred skills
Prompt-automation frameworks (Promptfoo, LangChain, Garak), vector search, LLM fine-tuning workflows, red-teaming certification
Technologies
Python, Bash, Jupyter, Promptfoo, LangChain, Garak
Responsibilities
Develop and run adversarial test suites for LLMs and image/video models; Craft multilingual prompts, jailbreaks, and escalation chains; Analyze outputs, triage failures, and write vulnerability reports; Contribute to internal tooling
Seniority
Mid-level (2-4 years experience)