TL;DR: 

I lead a project to research, design, and prototype an automated data analysis workflow for an internal mentorship program at a mid-sized agency.

Collaborating with a senior data scientist, I crafted an MVP data pipeline and dashboard which was human-centered, easy to use, and able to measure the success of 1:1 mentorship pairings at scale within the client’s organization.

Goal:

Research and design of a data pipeline and accompanying dashboard for a peer-to-peer mentorship program. 

Process:

Our project had three phases: discovery research, data pipeline creation, and iterative wireframing + prototype design:

1) Discovery Research

There were so many interviews to synthesize! 

2) Data Pipeline Development

 We created a set of human-centric qualitative KPIs to track mentor and mentee's growth, motivation, and satisfaction over time. 

To do this, we implemented statistical methods such as Weighted Averages, Harmonic Mean, and Dyadic Data Analysis into the survey workflow (diagramed below):

This way, we were able to quantify and measure the subjective experience of each mentor pairing.

3) Prototype Design

Next, I designed low-fidelity dashboard wireframes for feedback from stakeholders (shown below).

After iteration, we worked with a product designer for higher-fidelity wireframes. Finally, our data scientist coded this prototype into Google's Looker Studios.

Outcomes:

We applied quantitative rigor to an informal mentorship system, enabling the manager running the mentorship program to more accurately gauge the effectiveness and future trajectory of the program.

The mentorship program has continued to grow and mature since the completion of our project!