Case study 01 · 2025

Fifty lakh rupees were hiding in a spreadsheet nobody owned.

A licence register for a global pharma group became a platform leadership uses to make purchasing decisions.

RoleSole owner: discovery, design, build, rollout
OrganisationDishman Carbogen Amcis · MD's Office
StackFastAPI · Python · Next.js · AWS
Live atlicense.imdcal.com
License dashboard screenshot coming soon

The problem was memory, not software.

Across sites and functions, licences were tracked wherever the buyer happened to track things: a spreadsheet, an inbox, or a colleague's head. Nobody could reliably answer what the group owned, who used it, or what renewed next quarter.

Renewals were approved defensively, duplicates were bought in parallel, and unused seats quietly rolled over. I used stakeholder discovery to shape the schema, workflow, and technical boundaries before implementing the platform.

An internal tool nobody trusts is worse than the spreadsheet it replaced. Trust is a design requirement.
Key decisions

Three choices shaped the outcome.

01

Interview the process before writing the schema

I mapped how licences are actually requested, approved, and renewed across functions, including the informal steps nobody documents. The data model came out of those conversations, so the tool already spoke the users' language.

02

Make renewals the product, not a feature

The register itself saves nothing. Scheduled jobs escalate to the owner, then approver, then leadership, so a renewal becomes a decision made with time instead of an invoice discovered too late.

03

Own the rollout, not just the repository

I aligned with each business-side owner, migrated existing sheets, and stayed on the hook after launch. A platform handed over at deploy is a platform quietly abandoned by quarter's end.

Outcome

A quiet system with visible impact.

₹50Lavoided and recovered licence spend
1register replacing scattered spreadsheets
0renewals discovered after the fact
Livein production and owned end-to-end
KP.