Mercor

Artificial Intelligence

• Red-team conversational AI models and agents through jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Generate human data by annotating failures, classifying vulnerabilities, and flagging systemic risks
• Apply taxonomies, benchmarks, and playbooks to keep testing consistent
• Produce reproducible reports, datasets, and attack cases for customers
• Review AI outputs involving sensitive topics such as bias, misinformation, or harmful behaviors
• Uncover vulnerabilities automated tests miss
• Expand evaluation coverage and reduce production surprises
• Strengthen customer AI systems through adversarial testing

AI Safety Expert, English, Assamese

• Red-team conversational AI models and agents using jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Generate human data by annotating failures, classifying vulnerabilities, and flagging systemic risks
• Apply taxonomies, benchmarks, and playbooks to keep testing consistent
• Produce reproducible reports, datasets, and attack cases for customers
• Review AI outputs involving sensitive topics such as bias, misinformation, or harmful behaviors
• Uncover vulnerabilities that automated tests miss
• Expand evaluation coverage and reduce production surprises
• Help Mercor customers strengthen the safety, robustness, and trustworthiness of AI systems

AI Safety Expert – English, Bengali

Financial Planning and Analysis

• Complete an approximately 15-minute written intake about compliance practice
• Describe day-to-day workflows, time allocation, and workflows core to the role
• Explain how experienced practitioners divide the compliance field using insider perspectives
• Provide concrete examples of where AI gets compliance work wrong based on personal casework
• Identify invisible aspects of compliance work that outsiders and project designers tend to miss
• Contribute insights that shape the scope of future compliance engagements for AI labs
• Work on projects focused on training and enhancing AI systems

Financial Compliance Expert – CL Funnel

• Red-team conversational AI models and agents through jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Generate human data by annotating failures, classifying vulnerabilities, and flagging systemic risks
• Apply taxonomies, benchmarks, and playbooks to keep testing consistent
• Produce reproducible reports, datasets, and attack cases for customers
• Review AI outputs involving sensitive topics such as bias, misinformation, and harmful behaviors
• Help expand evaluation coverage and strengthen customer AI systems

AI Safety Expert – English, Gujarati

• Red-team conversational AI models and agents through jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Annotate failures, classify vulnerabilities, and flag systemic risks
• Apply taxonomies, benchmarks, and playbooks for consistent testing
• Produce reproducible reports, datasets, and attack cases for customers
• Review AI outputs involving sensitive topics such as bias, misinformation, or harmful behaviors
• Help Mercor customers make AI systems more robust, safe, and trustworthy

AI Safety Expert, English, Telugu

• Design adversarial prompts to stress-test frontier AI models
• Identify jailbreaks, unsafe behaviours, hallucinations, and policy failures
• Evaluate model robustness across misinformation, cyber, biosecurity, fraud, political content, and other sensitive domains
• Document vulnerabilities and contribute to safety benchmarking and red-teaming reports
• Collaborate with AI researchers to improve model alignment, robustness, and safety
• Complete projects on the contractor's own schedule
• Work on projects focused on training and enhancing AI systems

AI Safety Red Teamer

Japanese

• Open and check provided PDF tasks
• Confirm pages are in Japanese, legible, contain real content, and show no personal details
• Identify and bound every meaningful page region
• Assign component types and reading-order indices
• Link regions to associated figures or tables using parent component identifiers
• Transcribe all text exactly as shown, including Japanese scripts, furigana, and handwritten content
• Flag illegible source regions
• Capture language, document type, source, page dimensions, and table/formula/handwriting flags
• Review colleagues' annotation work end to end as an experienced annotator
• Build human-authored training data for document AI models

PDF Annotation, Transcription Expert – Japanese

Software Engineer

• Author domain-authentic materials science and engineering tasks based on prior work experience
• Build data rooms and grading criteria for objectively scorable tasks
• Run tasks against AI models
• Analyze model successes and failures
• Iterate tasks until they effectively discriminate model performance
• Collaborate with project leads and other domain experts to define quality standards

Materials & Engineering Specialist

General

• Write challenging single-turn prompts in radiological safety and health physics
• Label prompts as benign, dual-use, or adversarial
• Evaluate model responses against a defined policy standard
• Judge whether responses handled technical requests correctly
• Write reference answers and explain the technical reasoning
• Red-team frontier AI models for misuse potential in radiological safety scenarios
• Produce written rationales that non-specialists can follow
• Read and write about field-specific misuse scenarios for sustained periods

Health Physicist

• Red-team conversational AI models and agents using jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Generate human data by annotating failures, classifying vulnerabilities, and flagging systemic risks
• Apply taxonomies, benchmarks, and playbooks to keep testing consistent
• Produce reproducible reports, datasets, and attack cases
• Probe AI outputs involving bias, misinformation, and harmful behaviors
• Help Mercor’s AI-lab and enterprise customers train and enhance frontier AI models
• Expand evaluation coverage and identify vulnerabilities missed by automated tests

AI Safety Expert – English, Kannada

• Red-team conversational AI models and agents through jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Generate human data by annotating failures, classifying vulnerabilities, and flagging systemic risks
• Apply taxonomies, benchmarks, and playbooks to maintain consistent testing
• Produce reproducible reports, datasets, and attack cases for customers
• Review AI outputs involving sensitive topics such as bias, misinformation, and harmful behaviors
• Help expand evaluation coverage and reduce production surprises
• Collaborate with leading researchers and work on projects training and enhancing AI systems

AI Safety Expert, English – Danish

• Red-team conversational AI models and agents through jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Generate human data by annotating failures, classifying vulnerabilities, and flagging systemic risks
• Apply taxonomies, benchmarks, and playbooks to keep testing consistent
• Produce reproducible reports, datasets, and attack cases for customers
• Review AI outputs involving sensitive topics such as bias, misinformation, or harmful behaviors
• Help expand evaluation coverage and strengthen the safety, robustness, and trustworthiness of AI systems

AI Safety Expert – English, Punjabi

• Red-team conversational AI models and agents through jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Generate human data by annotating failures, classifying vulnerabilities, and flagging systemic risks
• Apply taxonomies, benchmarks, and playbooks to keep testing consistent
• Produce reproducible reports, datasets, and attack cases
• Review AI outputs involving sensitive topics such as bias, misinformation, and harmful behaviors
• Probe vulnerabilities that automated tests miss
• Expand evaluation coverage and reduce production surprises
• Strengthen customer AI systems through adversarial testing

• Red-team conversational AI models and agents through jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Generate human data by annotating failures, classifying vulnerabilities, and flagging systemic risks
• Apply taxonomies, benchmarks, and playbooks to keep testing consistent
• Produce reproducible reports, datasets, and attack cases for customers
• Review AI outputs involving sensitive topics such as bias, misinformation, and harmful behaviors
• Expand evaluation coverage and uncover vulnerabilities automated tests miss

Engineer

• Write challenging single-turn prompts in the propulsion and effects domain
• Label prompts as benign, dual-use, or adversarial
• Evaluate model responses against a defined policy standard
• Judge whether responses handled technical requests correctly
• Write reference answers and explain the technical reasoning behind them
• Distinguish routine professional questions from requests indicating potential misuse
• Sustain reading and writing about misuse scenarios in the field

AI Safety Expert, English, Assamese

AI Safety Expert – English, Bengali

Financial Compliance Expert – CL Funnel

AI Safety Expert – English, Gujarati

AI Safety Expert, English, Telugu

AI Safety Red Teamer

PDF Annotation, Transcription Expert – Japanese

Materials & Engineering Specialist

Health Physicist

AI Safety Expert – English, Kannada

AI Safety Expert, English – Danish

AI Safety Expert – English, Punjabi

AI Safety Expert – English, Punjabi

AI Safety Expert – English, Gujarati

Propulsion & Effects Engineer