Description
Good computer vision starts long before model training. If annotations are inconsistent, ambiguous, or poorly defined, the resulting model inherits those problems.
Angible is looking for an Associate Engineer who can make the labeling process technically rigorous. The position translates ML requirements into annotation rules, documents difficult edge cases, reviews output quality, measures consistency, and helps improve the connection between dataset quality and model performance.
Because Angible works with retail video and tracking problems, useful domain knowledge includes detection, multi-object tracking, Re-ID, frame-level labeling, QA processes, and dataset management.
This is therefore broader than repetitive manual labeling. It sits between the machine-learning team and annotation operations and gives the hire ownership over how data should be prepared and validated.
Please treat this as a curated summary and confirm the latest requirements on Angible's official job page.
https://www.angible.com/careers-associate-engineer-ai-labeling