Streamlining STP Data Collection & Dashboard Development

Hyundai Capital America · Analytics Internship

06/2026-08/2026

Project Context: Hyundai Capital America's Training & Content Design team tracked every agent training's completion and assessment results in a master STP (Speed-to-Proficiency) Tracker, but it was built entirely by hand. New trainings were retyped from Jira into the tracker one at a time, with no keys linking the two systems, which produced duplicate entries and structural errors across hundreds of logged trainings. I audited the tracker, rebuilt it around a relational data model, automated the reporting pipeline behind it, and delivered five dashboards giving leadership visibility they previously had to request through one person.

Business Question: How could HCA's Training & Content Design team replace a fully manual, error-prone STP reporting process with a connected system that leadership could explore on their own?

  • Fixed a fully manual, disconnected pipeline. Every Jira entry had to be retyped by hand into the STP Tracker with no keys linking the two systems, producing duplicate entries and structural errors across hundreds of trainings.
  • Cut monthly reporting effort by 60%. A Power Query workflow now cleans and standardizes Jira exports automatically instead of requiring manual recalculation every month.
  • Gave leadership self-service access. Every KPI or course-level question used to go through the one person who owned the tracker. Five interactive dashboards, with filters, parameters, and toggles, now let leadership explore assessment performance, trainer utilization, and class trends on their own.
  • Built it to last. Year-over-year and quarter-over-quarter comparisons were baked in from the start, so the dashboards stay useful as new training cycles roll in instead of needing to be rebuilt each time.
Methodology
  • Audit and data model: audited the existing STP Tracker for duplicate entries and structural errors, then built primary and foreign keys tied to each training's Jira ID to connect Jira, a new class-level layer, and the existing participant-level data for the first time.
  • Automation: used Power Query to clean and standardize Jira exports so new trainings, performance, and trainer hours auto-populate the tracker through formulas instead of manual retyping.
  • Dashboard build: built five Tableau dashboards (Monthly Overview, Assessment Overview, Class Details, Content Details, Trainer Tracker Overview) designed for the manager and leadership to maintain and explore independently.
Takeaways
  • The real deliverable was the connection between two systems that had never talked to each other, not the dashboards built on top of it.
  • Automation only works if it fits the existing workflow. Building around the manager's process, rather than replacing it outright, was what made the new system actually get adopted.
  • Presenting to executives meant translating a technical restructuring project into a business result: less manual work, faster reporting, and self-service access for leadership.