A portfolio helps demonstrate what a learner can do.
Each project should contain a title, business problem, dataset description and permitted usage, analytical questions, data-cleaning approach, methods and tools, code or analysis files, dashboard/visualisation, findings, recommendations, limitations, validation evidence and the learner’s explanation.
Projects must use lawful, appropriately licensed, synthetic or otherwise authorised datasets. Learners must not upload confidential client information, personal data or proprietary business data to public repositories or AI services without explicit authorisation. Public portfolio material must be reviewed for confidential information, credentials and personal data before release.
Five projects across the analytics toolchain.
Project 1 — Sales Performance Dashboard
Tools: Excel and Power Query
Clean a sales dataset, calculate KPIs, identify trends, compare categories, create a dashboard and prepare a short business summary.
Deliverable: Excel dashboard and written findings.
Project 2 — E-commerce Customer and Retention Analysis
Tools: SQL and PostgreSQL
Query transactional data, combine tables, calculate customer and order metrics, analyse repeat-purchase behaviour and prepare cohort comparisons.
Deliverable: SQL scripts, analytical outputs and findings report.
Project 3 — Customer Behaviour Analytics
Tools: Python, Pandas and NumPy
Load and inspect data, identify quality issues, clean and transform the dataset, calculate descriptive statistics, explore patterns and document limitations.
Deliverable: Jupyter notebook and analytical report.
Project 4 — Executive Business Intelligence Dashboard
Tools: Power BI, Power Query and DAX
Prepare data, create a data model, define measures, design report pages, add filters, validate metrics and prepare an executive summary.
Deliverable: Power BI report and management presentation.
Project 5 — AI-Augmented Analytics Assignment
Tools: An approved AI assistant alongside relevant analytical tools
Define the question, record assistance used, review generated output, test calculations/code, identify errors, document corrections and submit a validation log.
Deliverable: Analysis, validation log and reflective summary.
An end-to-end business problem.
Learners work through a business problem from initial definition to final presentation to an assessment panel.
Final deliverables: Business problem statement; dataset description and data-quality notes; analysis files and code; dashboard or visualisation; findings report; recommendations; presentation; and individual reflection.
Assessment principle: learners must be able to explain their work. A dashboard or code file without demonstrated understanding of its logic is insufficient.
Required stages
- Define the business problem and analytical questions.
- Inspect the dataset and document data-quality issues.
- Clean and prepare the data.
- Use SQL where database analysis is appropriate.
- Apply Python or spreadsheet analysis as required.
- Build a Power BI dashboard where appropriate.
- Validate calculations and findings.
- Identify meaningful insights and prepare evidence-based recommendations.
- Document limitations and assumptions.
- Present the results to an assessment panel.
Build work you can explain.
Ask the admissions team how projects and portfolio reviews are organised.