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
• Help expand evaluation coverage and strengthen customer AI systems

AI Safety Expert – English, Punjabi

• Red team conversational AI models and agents
• Test 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 actionable attack cases
• Uncover vulnerabilities automated tests miss
• Expand evaluation coverage and reduce production surprises
• Strengthen customer AI systems through adversarial testing

AI Safety Expert – English, Danish

Spanish

• Audit AI training data in Spanish rather than produce it
• Perform Multilingual Transcription Audits by listening to speech, reviewing annotator transcriptions, applying error codes, assigning pass/fail verdicts, and writing rationales
• Perform Forced Alignment Audits by verifying word-level segment timestamps and reference transcriptions
• Correct inaccurate boundaries, split or merge segments, add missing segments, and remove segments containing no word
• Tag every applicable error code and explain in plain language why each task passed or failed
• Call out critical errors first so failure reasons are clear
• Complete a short calibration phase before production work begins
• Work across Amazon Sonic audit collections
• Collaborate with leading AI researchers on projects training and enhancing AI systems

Sonic Audit Specialist, Spanish

Japanese

• Audit Multilingual Transcription data by listening to recordings and judging whether transcriptions accurately capture spoken content
• Apply fixed error codes, assign pass/fail verdicts, and write a short rationale for each judgment
• Audit Forced Alignment data by verifying word-level segment timestamps and reference transcriptions
• Correct inaccurate boundaries, split or merge segments, add missing segments, and remove segments containing no word
• Tag every applicable error code and explain in plain language why a task passed or failed
• Call out critical errors first so failure reasons are clear
• Complete calibration before beginning production work

Sonic Audit Specialist – Japanese

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

AI Safety Expert, English, Portuguese

• 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

Chemist

• Write challenging single-turn prompts in the chemistry and chemical safety 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 describing correct responses and the technical reasoning behind them
• Distinguish legitimate professional questions from requests indicating potential misuse
• Apply forensic and chemical casework expertise to assess AI safety behavior

Forensic, Toxicological Chemist

• Write challenging single-turn chemistry prompts labeled benign, dual-use, or adversarial
• Evaluate AI model responses against a defined policy standard
• Judge whether model responses handled technical requests correctly
• Write reference answers explaining what a correct response looks like and the technical reasoning behind it
• Red-team frontier AI models for chemistry misuse potential
• Apply practical route, scale-up, process chemistry, and chemical safety expertise to distinguish legitimate requests from dangerous ones
• Provide written rationales that non-specialists can follow
• Read and write about chemistry misuse scenarios while taking breaks or stepping away when needed

Synthetic / Process Chemist

Financial Planning and Analysis

• Design realistic, complex finance scenarios based on professional practice
• Build supporting materials including models, memos, filings, and data
• Author and evaluate finance tasks used to train AI agents
• Apply analytical and decision-making expertise in modeling, diligence, valuation, underwriting, forecasting, risk management, accounting, reporting, or related finance disciplines
• Construct complex problems requiring judgment rather than process high volumes
• Collaborate with leading researchers on AI training projects

Investment Finance Expert – CL Funnel

• Evaluate AI-generated valuation models, investment recommendations, industry analyses, and other analytical artifacts for quality and accuracy
• Compare AI-generated responses head-to-head based on analytical rigor, valuation soundness, and investment-recommendation quality
• Review financial analysis scenarios involving company valuation, capital structure, and portfolio decisions
• Provide expert feedback to help evaluate and improve AI systems
• Work on projects focused on training and enhancing frontier AI models

Financial Analyst, Investment Analyst

• Open and check provided PDF-page 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 indexes
• Link regions to their associated figures or tables using parent component identifiers
• Transcribe all text exactly as it appears, including Japanese scripts, furigana, and handwritten content
• Flag illegible source regions
• Capture language, document type, source, page dimensions, and table, formula, and handwriting flags
• Review colleagues' annotation work end to end
• Produce human-authored structural annotations and transcriptions for AI training data

PDF Annotation, Transcription Expert – Japanese

• Complete an approximately 15-minute written intake about compliance practice
• Describe day-to-day workflows, time allocation, and workflows central to the role
• Explain how experienced practitioners divide the compliance and financial-crime field
• Identify where AI gets compliance work wrong, using concrete casework examples
• Share overlooked or invisible aspects of compliance work
• Provide detailed, specific responses to shape the scope of future compliance engagements
• Work on projects focused on training and enhancing AI systems

Financial Compliance Expert – CL Funnel

• 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
• Probe AI outputs involving bias, misinformation, and harmful behaviors
• Expand evaluation coverage and identify vulnerabilities automated tests miss
• Contribute to training and enhancing frontier AI systems for Mercor’s customers

AI Safety Experts, English – Kannada

Product Designer

• Work on a research project for a leading AI Design company
• Work with an AI Design Agent to complete design tasks
• Apply professional design expertise to train and enhance AI systems
• Collaborate with leading researchers
• Complete projects on an individual schedule
• Deliver services according to project needs and performance expectations

Product Design Expert

• Evaluate the quality and accuracy of LLM-generated Spanish text across translation, summarization, dialogue, and grammar-correction use cases
• Edit, refine, or rewrite Spanish content to meet linguistic and cultural standards
• Annotate Spanish text for grammatical, syntactic, and semantic features
• Identify model errors and language gaps
• Collaborate with AI researchers to improve LLM performance
• Provide detailed linguistic feedback and analysis to guide ongoing model training
• Participate in projects training and enhancing AI systems