[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"job-entry-level-ai-jobs-united-states":3},{"Ques":4,"Slug":25,"Header":26,"job_category":42},{"title":5,"content":6},"Frequently Asked Questions",[7,10,13,16,19,22],{"A":8,"Q":9},"Yes. These roles are designed as entry points into production AI\u002FML workflows in the US, focusing on data labeling, RLHF evaluation, prompt evaluation, and QA evaluation that directly support LLM training pipelines and computer vision training.","Are these truly entry level AI jobs in the United States?",{"A":11,"Q":12},"Yes. Each role is explicitly marked Remote and uses the FULL_TIME employment type in the job metadata.","Are these roles remote and full-time?",{"A":14,"Q":15},"Common tasks include labeling text and images, applying annotation guidelines compliance checks, performing QA evaluation, handling edge cases, and creating training data quality improvements that impact model performance improvement.","What does an AI data annotation specialist do day to day?",{"A":17,"Q":18},"RLHF (Reinforcement Learning from Human Feedback) evaluation creates preference and scoring data by ranking model responses and applying rubrics, helping improve large language model evaluation outcomes like helpfulness, correctness, and safety.","What is RLHF evaluation and why is it included?",{"A":20,"Q":21},"Emphasize data labeling, RLHF, prompt evaluation, QA evaluation, annotation guidelines, training data quality, named entity recognition, computer vision annotation, content safety labeling, and LLM evaluation—plus consistent decision-making and clear documentation.","What skills should I emphasize when applying on Rex.zone?",{"A":23,"Q":24},"They can. Experience in training data quality, evaluation rubrics, and model behavior analysis often translates into roles such as QA lead, evaluation specialist, data operations analyst, or AI\u002FML operations roles supporting engineering teams.","Do these roles lead to other AI careers?","entry-level-ai-jobs-united-states",{"desc":27,"title":28,"content":29},"Entry level AI jobs in the United States focus on building and improving AI\u002FML systems through practical data operations: data labeling, RLHF evaluation, prompt evaluation, QA evaluation, and training data quality workflows for large language models and computer vision. On Rex.zone, these remote full-time roles support LLM training pipelines with annotation guidelines compliance, content safety labeling, and model performance improvement. You will collaborate with ops and engineering teams to produce high-quality datasets, track labeling accuracy, and help evaluate model outputs against rubrics. If you are exploring remote AI work in the US, this page helps you find and apply to roles aligned with real production AI workflows.","Entry Level AI Jobs in the United States",[30,33,36,39],{"h2":31,"desc":32},"Entry Level AI Data Annotation Specialist (United States, Remote)","Title: Entry Level AI Data Annotation Specialist (United States, Remote)\nDate: 25-02-2026\nCompany: Rexzone\nCountry: US\nRemote Type: Remote\nEmployment Type: FULL_TIME\nExperience Level: Mid-Senior\nIndustry: Technology\nJob Function: Engineering\nSkills: Data Labeling, RLHF, Prompt Evaluation, QA Evaluation, Training Data Quality, Annotation Guidelines, Named Entity Recognition, Content Safety Labeling, LLM Evaluation, Computer Vision Annotation\nSalary Currency: USD\nSalary Min: 63360\nSalary Max: 126720\nPay Period: YEAR\n\nYou will label and review text, image, and conversation datasets used in LLM training pipelines and computer vision training. You will apply annotation guidelines compliance checks, resolve edge cases, and document labeling decisions to improve training data quality.\n\nResponsibilities:\n- Perform data labeling across NLP and CV tasks (classification, ranking, segmentation, and NER)\n- Execute QA evaluation and inter-annotator agreement checks to reduce label noise\n- Support RLHF workflows by ranking responses and applying preference rubrics\n- Run prompt evaluation to assess instruction-following, factuality, and safety behaviors\n- Flag content safety labeling issues (policy, toxicity, sensitive content) with clear rationale\n- Track model performance improvement signals by auditing recurring model failure patterns\n\nQualifications:\n- Familiarity with large language model evaluation concepts (helpfulness, harmlessness, honesty)\n- Comfort following detailed rubrics and making consistent judgments at scale\n- Strong written communication for edge-case notes and escalation summaries\n- Basic understanding of NLP and computer vision annotation task types\n\nHow Success Is Measured:\n- Training data quality improvements (accuracy, consistency, and reduced rework)\n- Annotation throughput with high guideline adherence\n- Clear documentation that enables downstream engineering and model iteration",{"h2":34,"desc":35},"Entry Level AI\u002FML RLHF Evaluation Analyst (United States, Remote)","Title: Entry Level AI\u002FML RLHF Evaluation Analyst (United States, Remote)\nDate: 25-02-2026\nCompany: Rexzone\nCountry: US\nRemote Type: Remote\nEmployment Type: FULL_TIME\nExperience Level: Mid-Senior\nIndustry: Technology\nJob Function: Engineering\nSkills: RLHF, LLM Evaluation, Preference Ranking, Prompt Evaluation, QA Evaluation, Data Labeling, Safety Evaluation, Rubric Design, Model Behavior Analysis, Training Data Quality\nSalary Currency: USD\nSalary Min: 63360\nSalary Max: 126720\nPay Period: YEAR\n\nYou will evaluate LLM outputs to produce high-signal preference data for RLHF and supervised fine-tuning. This role is centered on prompt evaluation, rubric-based scoring, and QA evaluation to improve model behavior across helpfulness, correctness, and content safety.\n\nResponsibilities:\n- Rank multiple model responses and justify preference decisions using evaluation rubrics\n- Perform QA evaluation on peer work to maintain annotation guidelines compliance\n- Identify systematic model issues (hallucinations, refusals, instruction errors) and report trends\n- Support content safety labeling and safety evaluation for sensitive topics and policy areas\n- Contribute to training data quality by refining edge-case guidance and examples\n- Collaborate with ops\u002Fengineering partners to calibrate scoring and reduce ambiguity\n\nQualifications:\n- Strong critical reading and structured reasoning for consistent preference judgments\n- Ability to follow RLHF guidelines and maintain high inter-rater consistency\n- Experience interpreting evaluation criteria such as factuality, relevance, and safety\n\nHow Success Is Measured:\n- High agreement rates on preference ranking and rubric scoring\n- Reduction in rework through consistent QA evaluation\n- Actionable issue summaries that drive model performance improvement",{"h2":37,"desc":38},"Entry Level AI Content Safety Labeling Specialist (United States, Remote)","Title: Entry Level AI Content Safety Labeling Specialist (United States, Remote)\nDate: 25-02-2026\nCompany: Rexzone\nCountry: US\nRemote Type: Remote\nEmployment Type: FULL_TIME\nExperience Level: Mid-Senior\nIndustry: Technology\nJob Function: Engineering\nSkills: Content Safety Labeling, Policy Review, QA Evaluation, Data Labeling, LLM Evaluation, Prompt Evaluation, Risk Assessment, Annotation Guidelines, Safety Taxonomy, Incident Escalation\nSalary Currency: USD\nSalary Min: 63360\nSalary Max: 126720\nPay Period: YEAR\n\nYou will label and review content for safety categories to help train and evaluate AI systems. The work supports LLM training pipelines by producing reliable content safety labeling data and conducting QA evaluation against safety taxonomies.\n\nResponsibilities:\n- Apply safety taxonomies to label content across text and conversational datasets\n- Conduct QA evaluation to ensure annotation guidelines compliance and consistent policy use\n- Perform prompt evaluation to test model behavior on safety-sensitive instructions\n- Escalate high-risk items using incident escalation workflows with clear documentation\n- Help improve training data quality by updating examples, edge-case notes, and clarifications\n- Monitor recurring failure modes that affect model performance improvement (over-refusal, under-refusal)\n\nQualifications:\n- Strong judgment and attention to detail under detailed policy constraints\n- Ability to explain decisions concisely and consistently with documented rationale\n- Comfort working with sensitive content in a controlled, policy-driven environment\n\nHow Success Is Measured:\n- Consistent labeling accuracy and reduced guideline deviations\n- High-quality escalations with complete context\n- Measurable improvements in safety dataset reliability",{"h2":40,"desc":41},"Entry Level Computer Vision Annotation Specialist (United States, Remote)","Title: Entry Level Computer Vision Annotation Specialist (United States, Remote)\nDate: 25-02-2026\nCompany: Rexzone\nCountry: US\nRemote Type: Remote\nEmployment Type: FULL_TIME\nExperience Level: Mid-Senior\nIndustry: Technology\nJob Function: Engineering\nSkills: Computer Vision Annotation, Bounding Boxes, Polygon Segmentation, Keypoint Labeling, QA Evaluation, Annotation Guidelines, Dataset Curation, Data Labeling, Image Quality Review, Training Data Quality\nSalary Currency: USD\nSalary Min: 63360\nSalary Max: 126720\nPay Period: YEAR\n\nYou will create high-precision computer vision annotation for model training and evaluation, including bounding boxes, polygons, and keypoints. You will support dataset curation, quality review, and QA evaluation to maintain training data quality.\n\nResponsibilities:\n- Perform computer vision annotation: bounding boxes, polygon segmentation, and keypoint labeling\n- Follow annotation guidelines compliance checks and resolve ambiguous edge cases\n- Execute QA evaluation audits to reduce label noise and improve consistency\n- Support dataset curation by identifying low-quality images and annotation errors\n- Document annotation decisions to improve downstream model training outcomes\n\nQualifications:\n- Familiarity with common CV tasks and accuracy requirements for pixel-level labeling\n- Strong visual attention to detail and consistency under time-based throughput goals\n- Ability to learn new tooling and follow rubrics precisely\n\nHow Success Is Measured:\n- Accuracy of annotations and reduced correction rates\n- Stable throughput while maintaining training data quality\n- Clear notes that accelerate reviews and reduce ambiguity","AI Data Operations"]