Annotation Research Scientist (AI/Autonomous)

Annotation Research Scientist Role Summary

As an Annotation Research Scientist, you will research and implement algorithmic approaches for large-scale data annotation, quality assessment, and AI-assisted production workflows. This role requires both research judgment and engineering execution: you should be able to evaluate algorithmic directions, design reliable benchmarks, and collaborate with data platform, engineering, and delivery teams to bring research ideas into stable production workflows.

Key Responsibilities

• Research, evaluate, and implement automated or semi-automated annotation methods to improve data production efficiency and quality.
• Design algorithm-assisted workflows for complex data types, including annotation support, quality assessment, anomaly detection, and review optimization.
• Build evaluation frameworks, benchmarks, metrics, and error analysis methods to support data-driven algorithm selection.
• Collaborate with data platform, algorithm, engineering, and operations teams to turn research prototypes into reusable and maintainable production workflows.
• Analyze data quality issues and manual review bottlenecks, and propose practical approaches to improve efficiency and reliability.
• Stay up to date with advances in computer vision, weak/semi-supervised learning, automated annotation, and multimodal understanding, and validate relevant ideas in real scenarios.

Requirements

• Bachelor's degree or above in Computer Science, Artificial Intelligence, Electrical Engineering, Automation, Mathematics, or a related field; Master's or PhD preferred.
• 5+ years of experience in AI/CV/ML research or engineering. Experience with large-scale annotation, data quality, video understanding, multimodal data processing, or 3D perception is preferred.
• Strong foundation in computer vision or machine learning, with hands-on depth in at least one area such as video understanding, human pose estimation, 3D reconstruction, weak/semi-supervised learning, or automated annotation.
• Ability to independently investigate, reproduce, adapt, and evaluate algorithmic approaches, beyond simply reading papers.
• Strong engineering mindset with the ability to integrate algorithm modules into real data workflows while considering stability, cost, maintainability, and scalability.
• Proficiency in Python and common deep learning frameworks, with solid experiment management, data analysis, and result interpretation skills.
• Strong independent research judgment, ownership, curiosity, and resilience, with the ability to drive validation and delivery in an early-stage startup environment.

Nice to Have

• Experience deploying algorithms in large-scale video, multimodal, or spatial data production systems.
• Experience in automated annotation, active learning, weak/semi-supervised learning, data quality evaluation, or manual review cost reduction.
• Experience with robotics, embodied AI, industrial vision, AR/VR, or complex sensor data.
• Publications, open-source projects, or strong technical output in CVPR / ICCV / ECCV / NeurIPS / ICLR / ICML or related venues.
• Experience in AI data services, annotation platforms, AI infrastructure companies, or research labs.

Regret to inform that only shortlisted candidates will be notified. www.dadaconsultants.com EA Registration Number: R1878287 Business Registration Number: 201735941W.