Mastercard Advisors · May 2026 – Jul 2026
Cutting quarterly reporting from 120 minutes to 5
Every quarter, the team I worked in builds review files for India's #1 card issuer. Each file took about two hours of manual spreadsheet work, across 10 file types. I built a pipeline in Python and VBScript that produces each file in 5 minutes with no errors, and the team adopted it as its standard way of working.
- Company
- Mastercard AdvisorsMastercard's consulting and analytics arm
- My role
- Associate Consultant Intern, Performance Analytics
- When
- May 2026 – Jul 2026
- Tools
- Python · SQL · VBScript · Microsoft Excel
- Source120 → 5 min per file
- 20–25
- 10
The client stays anonymous, and there are no screenshots or client data on this page, by design.
The problem
QBR production was manual. Each file took about 120 minutes of pivot-table work, there were 10 file types, and the total came to 20+ hours of analyst time every quarter.
Manual work is also where mistakes creep in, in numbers a client relies on. And because it was manual, taking the same reports to more clients would have meant more hours, not fewer.
What I did
- Built a parameter-driven pipeline in Python and VBScript. You choose the inputs, run it, and get the finished file, with no copy-pasting or hand-built pivots.
- Made the metrics modular. SME and market metrics are separate building blocks, so adding or changing a metric doesn't mean rewriting the pipeline.
- Made the output deterministic and auditable. The same inputs always produce the same file, and every number can be traced back to its source.
- Designed it to scale beyond this client, as a reporting framework the team can reuse for other banking clients.
- Handed it over to the team, who adopted it as their standard way of producing QBRs.
What I'd do next
Roll the pipeline out to other bank clients, which is what it was designed for, and add automatic checks that flag unusual quarter-on-quarter movements before a person reviews the file.
Key terms (5): new to the jargon?
- QBR (quarterly business review)
- A recurring report that shows a client how its business performed over the quarter and where to act next.
- Pivot table
- A spreadsheet tool for summarising large tables. Building many of them by hand every quarter is slow and easy to get wrong.
- Parameter-driven
- You set a few inputs (such as which quarter and which file) and the pipeline produces the output. Nothing is edited by hand.
- Deterministic and auditable
- The same inputs always produce exactly the same file, and every number can be traced back to where it came from.
- SME
- Small and medium-sized enterprises, one of the customer segments the reports cover.
What's next