Data Science · Analytics Engineering · Reporting Automation
CompletedEST Assessment Data Engineering
Automated pipelines and dashboards for multi-country exam data
Analytical workflows for large-scale EST exam datasets covering students across Egypt, Jordan, Saudi Arabia and the UAE: Python pipelines, statistics, Dash dashboards and automated reports.
- Organization
- York Press
- Role
- Data Analyst
- Timeframe
- July 2024 – October 2025
The problem
Large exam datasets across several countries, subjects and test versions were reported through manual workflows that were slow and prone to transcription errors. Different test versions also made like-for-like comparison difficult.
The objective
Automate processing and reporting, and make performance comparable across subjects, skills, countries and test versions.
My role
- Built automated Python pipelines to process and visualise the exam datasets.
- Developed Dash dashboards tracking student performance against predefined skill sets.
- Designed CSV reporting templates for schools and stakeholders.
- Partnered on the design and deployment of school performance reporting tools.
Technical approach
- Python (pandas, NumPy) pipelines ingest, clean and reshape raw exam results.
- Skill-, subject- and country-level statistics (mean, median and standard deviation) across Mathematics, Literacy, Biology and Chemistry.
- Test items are mapped to target skills per version, enabling comparison across otherwise non-equivalent test forms.
- Manual reporting replaced with automated Excel, Power BI and Python processes.
Key features
Automated pipelines
Repeatable processing from raw results to report-ready tables.
Skill mapping
Item-to-skill mapping per test version.
Performance dashboards
Dash dashboards by subject, skill, country and version.
School reports
CSV templates summarising scores, version distribution and skill metrics.
Interactive demo
SYNTHETIC DATAAll figures are randomly generated for demonstration. They do not describe real students, schools or results.
Students
1,200
synthetic records
Mean
62%
Median
62%
Std deviation
14.0
percentage points
Pass rate
81%
at 50% pass mark
Score distribution
Mathematics · number of students per 10-point band
- Band below pass mark
- Band at or above pass mark
Mean score by country
Mathematics · version All
Mean score by skill
Items mapped to target skills
Test-version comparison
Mathematics · all countries
Outcome
- Manual reporting workflows replaced with automated processes, cutting turnaround time and eliminating recurring transcription errors.
- Like-for-like comparison across test versions through skill mapping.
No student records or exam content are shown. The dashboard below uses synthetic data.