Remote Red Team: Explosives & Energetic Materials Expert

Mercor; San Francisco (CA); À distance; GBP 67 000 - 77 000; Plein temps

Résumé du poste

Mercor seeks an Energetic Materials Expert for Redteaming on a remote contract basis. The role involves crafting challenging prompts about explosives and energetic materials, and evaluating AI model responses under a defined policy framework.

Successful candidates will label prompts, draft reference answers with technical reasoning, and identify misuse while ensuring safe, responsible AI interactions. Prior domain expertise is essential for timely, independent delivery.

Qualifications

• Must have hazardous device expertise with render-safe procedures and device assessment.
• Specialty materials chemistry expertise in pyrotechnics or propulsion.
• Federal certification in materials with incident analysis or forensic capabilities.
• Background in munitions propulsion design and effects engineering.
• Experience in specialty materials handling and blasting operations.

Responsabilités

• Write challenging single-turn prompts on explosives/energetic materials and label as benign, dual-use, or adversarial.
• Evaluate model outputs against policy standards and judge correctness.
• Create reference answers with correct responses and technical reasoning.
• Identify misuse potential in requests and ensure safe model refusals for dangerous queries.
• Work independently and asynchronously to meet deadlines and improve AI model performance.

Connaissances

Hazardous device specialist

Specialty materials chemist

Federal certified materials specialist

Propulsion & effects engineer

Commercial blasting engineer

Description du poste

Mercor seeks an Energetic Materials Expert for Redteaming on a remote contract basis. The role involves crafting challenging prompts about explosives and energetic materials, and evaluating AI model responses under a defined policy framework.

Successful candidates will label prompts, draft reference answers with technical reasoning, and identify misuse while ensuring safe, responsible AI interactions. Prior domain expertise is essential for timely, independent delivery.