Aelven

Aelven (legal entity Aelven Inc.) is a human data company that builds the datasets used to train AI models, describing itself as the connective layer between humans and AI. Rather than scraping existing text, it captures expert work and real-world context: judgment, correction, and intent, the parts of how a person actually reasons through a task that raw data does not record. It runs custom data programs for AI labs alongside its own proactive collection initiatives, and builds agent-native tooling and quality systems so contributors produce consistent, structured data at scale. Stated focus areas include measuring long-horizon agent capabilities and cross-modal context, meaning the harder evaluation problems rather than simple labeling. For workers, Aelven hires AI Model Training Specialists on a remote contract basis, currently US-only, paying $20 to $30 per hour depending on project type, quality, and experience, with flexible hours up to 40 per week depending on project availability. The work is reviewing, labeling, and evaluating data used to train AI systems: judging whether tasks, workflows, or outputs follow written instructions, comparing examples to decide which better meets project criteria, and identifying errors, inconsistencies, missing information, or unclear reasoning. Projects span language tasks, workflow data, robotics-related data, structured evaluations, and AI-generated content. Notably, no degree and no prior AI or annotation experience are required, and the qualification task is paid, which is worth knowing because unpaid qualification and training time is a common grievance across this category. Aelven does not publicly disclose its founders, headquarters, founding year, headcount, or funding, so treat its corporate details as unknown. It is unrelated to Aiven (data infrastructure) or Alven (a venture capital firm).

Aelven's main contributor role is AI Model Training Specialist, and it is deliberately open to beginners. There is no degree requirement and no prior AI or data annotation experience needed. What the listing actually asks for is critical thinking, attention to detail, strong reading comprehension, and fluent English, since the work is judging whether AI outputs follow written instructions and spotting where reasoning breaks down. Practical requirements: you must be based in the United States, at least 18 years old, have a reliable internet connection and a laptop or desktop computer, complete a 30-minute recorded general assessment, and complete a qualification task, which Aelven pays for. Availability to start promptly is preferred. Day to day you review, label, and evaluate training data across language tasks, workflow data, robotics-related data, structured evaluations, and AI-generated content, comparing examples against project criteria and flagging errors, inconsistencies, missing information, or unclear reasoning. Pay is $20 to $30 per hour depending on project type, quality, and experience, on a contract basis with flexible hours up to 40 per week, subject to project availability rather than guaranteed. Aelven also posts an open application for full-time US remote company roles. Apply at aelven.ai/careers; listings are also carried on Glassdoor and Remotive.