The Data Mine, Purdue University · Corporate partner: Johnson & Johnson
In progress
An ongoing Purdue Data Mine team project, with Johnson & Johnson as corporate partner, focused on improving a scheduling and coordination workflow that is handled manually today. On the Package Generation team, my work so far has centred on defining the problem and specifying how a solution would be built.
Johnson & Johnson is the corporate partner for this Purdue Data Mine project. I contribute as a Purdue graduate researcher and am not employed by Johnson & Johnson.
QSTEPS AI-Enablement is a team project run through The Data Mine at Purdue University, with Johnson & Johnson as the corporate partner and J&J mentors guiding the work.
The project is focused on improving a scheduling and coordination workflow that is currently carried out manually. The work is organized across teams, and I'm part of the Package Generation team.
The project is ongoing. This page covers my contributions to date and will grow as the work develops.
At the centre of the project is a scheduling and coordination workflow that is handled manually today.
Before anything gets built, the team needs a clear, shared answer to a few basic questions. Getting those right early is what keeps the rest of the work pointed at the real problem — and that is exactly what the PRD is for.
The questions to answer
What exactly is the problem with the current workflow?
Who is the primary user we are solving for?
What is in scope — and what is deliberately out of scope?
How will we know the workflow has actually improved?
I'm a Graduate Data Science Researcher on the Package Generation team.
My focus has been the documentation that defines and plans the work — the Product Requirements Document (PRD) and the Technical Specification Document (TSD) — alongside day-to-day collaboration with my team in GitHub and Jupyter Notebook.
What to solve, and for whom.
Sections I wrote
Problem statement — what is not working in the current manual workflow, stated plainly enough for the whole team to align on.
Primary-user definition — who the solution is for, so decisions can be checked against a real person's needs.
Project scope — what the project covers and what it deliberately leaves out.
Success metrics — how the team will judge whether the workflow has improved.
How it would be built, and what happens when things go wrong.
Sections I wrote
Implementation by phase — how the work is broken into stages that can be delivered one at a time.
Data and systems touched — which data and systems the solution would need to interact with.
Error handling and fallback — what should happen when a step fails, so the workflow can keep moving.
All of this happens as part of the Package Generation team, working alongside teammates and J&J mentors rather than in isolation.
GitHub
Collaborating with the team on project work.
Jupyter Notebook
Used for project work during the engagement.
Status — In progress
Work in progress. This page will be updated as the project develops.
