Data Analyst [AQ-23432]

About the role

We are seeking a detail-oriented Data Analyst with Azure storage experience to support OCR processing and the management of large document and text datasets used for AI model training. This role focuses on monitoring existing OCR pipelines, reviewing generated data, maintaining accurate metadata and inventories, and managing Azure uploads and downloads. The analyst will work closely with engineers and applied scientists to identify issues, coordinate reprocessing, and ensure datasets are complete, organized, and ready for downstream use.

Responsibilities

OCR Pipeline Monitoring

• Operate existing OCR workflows using established tools and procedures.
• Monitor processing queues, job status, completion rates, and failed or stalled batches.
• Review logs and processing reports to identify exceptions; perform documented recovery steps and escalate technical issues to engineers.
• Maintain batch tracking records and coordinate reruns, priorities, and outstanding processing tasks.

Data Quality and Review

• Review OCR outputs against source documents using sampling and established quality criteria.
• Identify missing pages, empty outputs, duplicate records, unreadable text, formatting problems, and incomplete content.
• Run existing cleaning and validation tools; review results and flag questionable changes for further investigation.
• Maintain issue logs, categorize recurring errors, and track corrections through resolution.
• Summarize quality trends and recommend improvements to review procedures and data handling.

Dataset and Metadata Management