Headline

CXQO Associate-Machine Learning Data Associate at Amazon Go - (JWO Tech) |(AWS)

About

1. Data Annotation & Labeling: o Manually label images, audio, text, or video data. o Tag and categorize data accurately according to predefined guidelines. 2. Data Quality Assurance: o Review and validate labeled data for accuracy and consistency. o Provide feedback to improve data annotation guidelines and processes. 3. Tool Usage: o Use specialized tools and platforms to annotate and process data. o Work with internal software systems and possibly provide feedback for improvements. 4. Collaboration: o Coordinate with ML engineers, data scientists, and project managers. o Communicate issues related to data quality or process inefficiencies. 5. Reporting & Documentation: o Maintain documentation of work and progress. o Report on the quality and quantity of data processed. • Strong attention to detail. • Good understanding of data handling and labeling techniques. • Basic understanding of machine learning concepts (optional but beneficial). • Familiarity with annotation tools (e.g., Labelbox, CVAT, Amazon SageMaker Ground Truth). • Ability to follow guidelines and adapt to changing requirements. • Video/Image Annotation: Review and label hours of in-store camera footage according to strict guidelines. Identify people, products, and actions (who picked up or returned each item) to create training data for the ML models. Draw bounding boxes or tags on video frames using specialized annotation software. • Sensor Data Cross-Check: Verify that visual labels match other sensor inputs (e.g. shelf-weight triggers or RFID scans), involves confirming whether an item-removed event in the system aligns with the video annotation. This is part of sensor fusion validation (integrating camera and weight data). • Customer/Behavior Tracking: Help track each shopper throughout the store (person detection and multi-camera tracking) and note interactions. For example, flag occlusions or tangled states and ensure people are continuously identified. Annotate subtle behaviors (e.g. pose changes or hand movements) that signal a product pick-up or put-back. • Quality & Throughput: Hit high accuracy and speed targets,to achieve required quality (typically >95% accuracy) and output targets. • Process Improvement: Identify and suggest ways to improve annotation workflows and tools. Participate in process improvement to increase the quality, efficiency, and accuracy. • Issue Reporting & Collaboration: Work closely with managers and engineers. Flagging unclear cases or software errors during annotation and getting guidance from data scientists or developers.

Location

New Delhi

Work experience

CXQO Associate -Machine Learning Data Associate at Amazon, Jun 2021 - Present · 4 yrs 11 mos, New Delhi, Delhi, India, Working on Vision algorithm , AI , Computer Vision , Deep learning Algorithms, Sensor Fusion, Just Walkout Technology

ENTRY CLEARANCE ASSISTANT at British High Commission in India, Mar 2020 - Aug 2020 · 6 mos, New Delhi, Delhi, India

LE02 at High Commission of Canada in India, Oct 2018 - Sep 2019 · 1 yr, New Delhi, Delhi, India

Education

Bachelor of Technology - BTech, Electrical, Electronics and Communications Engineering, Vivekanad Institute of Technology and Science

Intermediate, Mathematics and Computer Science, North East Railway Senior Secondary School