Mercor

Voice Actor

• Record high-quality voice samples across conversational, narrative, instructional, and other scripts
• Deliver clear, natural, expressive speech with control over tone, pacing, and pronunciation
• Maintain consistency in voice, accent, and delivery across recording sessions
• Follow recording guidelines for environment, microphone setup, and file formatting
• Perform multiple takes with variation in emotion, emphasis, and style when required
• Contribute recordings to training and evaluation of next-generation text-to-speech systems
• Recordings will be used exclusively for the client's internal CX AI Agent and may be voice-cloned for that use case

Voice Actor, CX Agent Voice Cloning – Standard French

Artificial Intelligence

• 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, and harmful behaviors
• Uncover vulnerabilities that automated tests miss
• Expand evaluation coverage and reduce production surprises
• Strengthen customer AI systems through adversarial testing

AI Safety Experts, English, Kannada

Associate

• Build a working financial model of an assigned consumer / leisure transaction from a synthetic deal document pack
• Deliver an editable Excel workbook with live formulas, source references, and stated assumptions
• Peer-review two anonymized models from other contractors against a fixed set of criteria
• Provide a short written rationale for each peer review and complete a review scorecard
• Work on projects focused on training and enhancing AI systems

Private Equity Associate – Consumer & Leisure

• Build a working financial model of an assigned consumer / leisure transaction from a synthetic deal document pack
• Deliver an editable Excel workbook with live formulas, source references, and stated assumptions
• Peer-review two anonymized models from other contractors against a fixed set of criteria
• Provide a short written rationale for each peer review
• Complete a review scorecard
• Complete approximately 20–25 hours of work over 1.5–2 weeks
• Work on projects focused on training and enhancing AI systems

• 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 to keep testing consistent
• Produce reproducible reports, datasets, and attack cases
• Probe AI outputs involving bias, misinformation, and harmful behaviors
• Uncover vulnerabilities automated tests miss
• Expand evaluation coverage and reduce production surprises
• Strengthen customer AI systems through human data-driven red teaming

AI Safety Expert – English, Kannada

General

• Write challenging single-turn prompts in radiological safety, labelled benign, dual-use, and adversarial
• Evaluate model responses against a defined policy standard
• Judge whether responses handled radiological safety requests correctly
• Write reference answers and explain the technical reasoning behind them
• Create prompts that distinguish legitimate professional questions from dangerous misuse scenarios
• Provide written rationales that non-specialists can follow
• Read and write about radiological-safety misuse scenarios for sustained periods
• Scope the work appropriately where classified, export-controlled, NDA-covered, or prepublication-restricted obligations exist

Radiation Safety Officer

• Write expert-level prompts in Turkish across sensitive subject areas
• Apply structured guidelines to classify prompts and conversations
• Flag adversarial phrasings and escalation patterns
• Document the reasoning behind judgments
• Evaluate and strengthen how advanced AI models handle sensitive topics in Turkish
• Work on projects focused on training and enhancing AI systems

Bilingual Turkish Generalist Expert – AI Safety

Chemist

• Write challenging single-turn prompts in chemistry and chemical safety, labelled as benign, dual-use, or adversarial
• Evaluate model responses against a defined policy standard
• Judge whether each response was handled correctly
• Write reference answers describing what a correct response looks like and the technical reasoning behind it
• Red-team frontier AI models on their ability to assess technical-request misuse potential
• Draw reliable distinctions between legitimate process-chemistry questions and genuinely dangerous requests

Synthetic, Process Chemist

• 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 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
• Identify vulnerabilities that automated tests miss
• Expand evaluation coverage and reduce production surprises
• Strengthen customer AI systems through adversarial testing

AI Safety Expert – English, Gujarati

• 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
• Work on projects focused on training and enhancing AI systems

AI Safety Red Teamer

• Record and produce French-language audio transcripts for an AI lab's model training and audio-data improvement
• Use a neutral-to-upbeat customer-service voice with clear enunciation and consistent pacing
• Follow written punctuation and transcript instructions precisely
• Edit transcripts as needed to reflect natural pacing
• Complete approximately 2–10 hours of work per week for roughly 1–2 weeks
• Collaborate with leading AI researchers through Mercor projects
• Work remotely on a self-directed schedule

French Voice Actor

• Write challenging single-turn prompts in chemistry and chemical safety, labeled as benign, dual-use, or adversarial
• Evaluate model responses against a defined policy standard
• Judge whether responses were handled correctly
• Write reference answers describing correct responses and the technical reasoning behind them
• Red-team frontier AI models to assess whether they identify misuse potential in technical requests
• Distinguish legitimate process chemistry questions from requests seeking dangerous information
• Sustain reading and writing about misuse scenarios in the chemistry domain

• 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
• Help Mercor's customers train and enhance frontier AI models
• Collaborate with leading researchers and contribute to safer, more robust, and trustworthy AI systems

AI Safety Expert, English – Bengali

• Write challenging single-turn prompts in chemistry and chemical safety, labelled as benign, dual-use, or adversarial
• Evaluate model responses against a defined policy standard
• Judge whether responses handled technical requests correctly
• Write reference answers explaining what a correct response looks like and the technical reasoning behind it
• Create prompts that distinguish routine process-chemistry questions from requests involving potential misuse
• Provide written rationales understandable to non-specialists
• Read and write about misuse scenarios in chemistry for sustained periods

Synthetic – Process Chemist

• 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
• Uncover vulnerabilities that automated tests miss
• Expand evaluation coverage and reduce production surprises
• Work on projects focused on training and enhancing AI systems

Voice Actor, CX Agent Voice Cloning – Standard French

AI Safety Experts, English, Kannada

Private Equity Associate – Consumer & Leisure

Private Equity Associate – Consumer & Leisure

AI Safety Expert – English, Kannada

Radiation Safety Officer

Bilingual Turkish Generalist Expert – AI Safety

Synthetic, Process Chemist

AI Safety Expert – English, Gujarati

AI Safety Red Teamer

French Voice Actor

Synthetic, Process Chemist

AI Safety Expert, English – Bengali

Synthetic – Process Chemist

AI Safety Expert – English, Assamese