Jobs at Welo Global

Evaluate AI model outputs, conduct audit-based reviews, identify patterns and failure modes, apply and improve evaluation guidelines, and document actionable findings. The role requires native-level Marathi, strong English, structured writing, GenAI safety knowledge, critical thinking, and comfort making judgment calls in ambiguous situations. Associates collaborate with leads and stakeholders, participate in calibrations, and help refine quality standards. This is a full-time, one-year, 100% onsite contract role in specified U.S. cities.

Evaluate AI model outputs, audit data and model behavior, identify edge cases and failure modes, apply and improve evaluation guidelines, communicate performance trends, participate in calibration discussions, and document actionable findings. The role requires native-level Tamil, strong English and writing skills, critical judgment, attention to detail, and familiarity with GenAI safety considerations. This is a full-time, one-year, 100% onsite contract position.

Lead and manage Croatian linguistic testing for a long-term onsite project: ensure language quality, coordinate and train testers, handle client communication, manage linguistic assets, execute quality plans, and deliver timely, accurate localized content.

Evaluate AI outputs, perform audit-based reviews, identify failure modes and trends, apply and refine guidelines, communicate findings, participate in calibrations and stakeholder syncs, and help improve evaluation frameworks for human-in-the-loop AI.

Evaluate AI outputs, audit model behavior, apply and refine labeling guidelines, identify failure modes and trends, document findings, and collaborate with cross-functional teams to improve model evaluation and data quality.

Audit and review data labeling outputs, provide actionable feedback, train and support labeling teams, monitor workflows, identify recurring issues and edge cases, participate in calibrations and stakeholder syncs, and help improve guidelines and processes in human-in-the-loop AI workflows.

Audit and ensure quality of data labeling outputs, coach and onboard Data Labeling Associates, provide actionable feedback, monitor workflows, identify patterns and edge cases, and help improve guidelines and processes in human-in-the-loop AI workflows.

Audit and validate data labeling outputs, provide structured feedback and hands-on training to annotators, monitor workflows and escalate issues, help apply and improve guidelines, and identify patterns in human and model outputs to maintain high-quality AI training data.

Audit and ensure quality of labeled data by reviewing DLA outputs, providing feedback and coaching, supporting onboarding and training, monitoring workflows, identifying recurring issues and edge cases, and contributing to guideline and process improvements in human-in-the-loop AI workflows.

Evaluate AI model outputs, perform audit-style reviews to find patterns and failure modes, apply and refine labeling/evaluation guidelines, document findings, communicate insights to leads and stakeholders, and participate in team calibrations to improve model evaluation and data quality.

Evaluate AI model outputs, perform audit-based reviews, identify failure modes and trends, apply and improve guidelines, document findings, and collaborate with leads and cross-functional teams to refine quality standards.

Evaluate AI model outputs, perform audit-based reviews, apply and refine guidelines, identify performance trends and failure modes, document findings, and collaborate with cross-functional teams to improve model evaluation and data quality.

Audit and monitor data labeling outputs, provide actionable feedback, support onboarding and training of annotators, enforce and improve guidelines, escalate workflow issues, and document findings to maintain high-quality human-in-the-loop AI data processes.

Audit and ensure quality of data labeling work for ML using human-in-the-loop processes. Provide feedback, train and support annotators, monitor workflows, identify recurring issues, and help improve guidelines and processes.

Perform on-site data annotation and QA for natural language datasets, test AI products, analyze data for consistency and relevance, report anomalies, support ML model updates, and prepare concise reports on data quality and model performance.

Evaluate AI outputs, perform audit-based reviews, identify patterns and failure modes, apply and improve annotation/evaluation guidelines, document findings, and communicate insights to cross-functional teams to improve model performance and evaluation frameworks.

Evaluate AI model outputs, perform audit-based reviews to identify patterns and failure modes, apply and refine guidelines, document findings, and collaborate with leads and cross-functional teams to improve model evaluation and data quality.

Operate I2RT YAM bimanual robotic arms for scripted data-collection tasks, including teleoperation and human-robot interactions. Follow SOPs, coordinate with a station partner, maintain safety and data-handling standards, log trial results and anomalies, and assist with resets and retries. Training is provided, with emphasis on reliable, precise, repetitive work and clear communication.

Lead onsite execution of a robotics data collection pilot using bimanual robotic arms. Coach teleoperation and HRI operators, enforce SOP, safety, and quality standards, monitor live sessions and data quality, troubleshoot issues, document blockers, and coordinate with client stakeholders. Contribute to improving SOPs and checklists while maintaining station readiness and consistent execution.

Own Uzbek Cyrillic language quality for UI, marketing, product, legal, informational, and support content. Serve as the primary client-facing linguistic expert, interpret style guides and quality frameworks, manage arbitrations and escalations, coach reviewers, facilitate alignment sessions, track quality trends, and convert recurring issues into style guide improvements. The role is remote freelance work of approximately 24 hours per month for candidates in approved countries.