Data Engineering Researcher

Main job duties:

1. Multi-Modal Sensor Data Pipeline & Processing:

• Ingest, clean, and align high-frequency time-series telemetry from AIRmotion’s hardware stack, including 28-node IMU spatial vectors, full-body capacitive sensing matrices, and pneumatic pressure logs.
• Address real-world sensor data anomalies: handle missing sensor frames, time-drift across IMU nodes, dropped packets, and zero-calibration baseline offsets.
• Build standardized raw and analytical storage layers (Parquet/CSV) partitioned by user session, experimental posture regions, and actuator state events.

2. Feature Engineering & Dataset Preparation for Physical AI:

• Develop repeatable Python cleaning and feature-generation scripts (hosted via AWS Lambda / Step Functions) to output training-ready datasets for algorithm engineers.
• Extract spatial-temporal features supportive of dynamic sleep posture classification (supine, lateral) and body mass/morphology distribution modeling.
• Maintain session manifest files, dataset versioning, and feature contracts for offline model training and shape-servoing algorithm validation.

3. Data Cataloging, Querying & AWS Infrastructure:

• Catalog raw and transformed S3 sensor logs using AWS Glue Data Catalog, enabling team-wide analytical querying via Amazon Athena SQL.
• Maintain clear data dictionaries documenting physical sensor units (e.g., tilt angles, pressure values, capacitive presence), timestamp offsets, and hardware board revisions.

4. Data Quality & System Observability

• Implement automated data quality checks (timestamp monotonicity, frame completeness, physical sanity bounds, export validation) for every recorded trial session.
• Set up automated error tracing and log monitoring at the pipeline level using AWS CloudWatch.
• Collaborate closely with firmware, backend, and AI algorithm engineers to validate feature specs and push validated pipeline logic into production workflows.

Entry requirement

• Degree (or equivalent background) in Computer Science, Data Engineering, Software Engineering, Biomedical Engineering, or a related field.
• Hands-on proficiency with Python for data manipulation (e.g., Pandas, NumPy, file I/O) and AWS SDK (Boto3) to build reproducible automation scripts.
• Strong SQL foundation (filtering, aggregations, window functions, joins) with experience or willingness to query S3 via Amazon Athena.
• Basic familiarity with time-series data structures (JSON, CSV, Parquet), AWS S3, and IAM permission principles.
• Good documentation habits to communicate schema assumptions clearly with backend and algorithm teams without relying solely on oral syncs.

Job Type: Full-time

Pay: $18,000.00 - $23,000.00 per month

Benefits:

• Additional leave
• Paid sick time

Work Location: In person