Milestone 1

Deadline: Midnight of June 17

Introduction

As we continue to build on the complexity of the data pipelines over the course of the semester, they should always be governed by the business questions we are interested in answering. To this end, these business decisions will compound over the course of the semester, and future milestones should still meet all previous milestone’s business decisions. This may be as simple as not touching what was done before, and everything still works. But if you need to make architectural changes, it may require revisiting older milestones to ensure they still operate correctly. In general, you should plan on each business decision mapping to a single dashboard. You could reuse components of existing dashboards, but it should be clear in Grafana what business decision each dashboard maps to. Keep in mind that for these assessments, what matters is the accuracy and degree to which the provided dashboard helps inform the business decision. The technical means by which you compiled and created that dashboard is not considered here, but should be included in your documentation.

The Business Questions

For this first milestone, you will have two business questions that you are trying to answer and provide insight to:

Decision 1

Prague’s metro authority is preparing for a seasonal uptick in ridership, spurred by upcoming festivals, university terms restarting, and a gradual return to office work in the city center. With only a limited number of trains available for active rotation due to ongoing maintenance and staffing constraints, operations managers need to make data-driven decisions about where those trains will make the biggest impact. Understanding which lines experience the greatest demand—and how that demand shifts throughout the day—is essential to preventing overcrowding during rush hours and ensuring consistent service in underserved areas. A clear picture of usage patterns by line and hour will help planners dynamically allocate rolling stock to match actual ridership needs, rather than relying on outdated schedules or intuition.

Decision 2

While Prague’s metro system is known for its speed and efficiency, recent anecdotal complaints suggest that travel times may be creeping upward. Whether it is lingering platform congestion, increased transfer activity, or subtle slowdowns in service frequency, the operations team needs a baseline to validate against whether these perceptions reflect a measurable trend. By analyzing ride durations across different days of the week, planners can spot patterns in how long riders are spending in the system. Understanding the distribution of ride durations—and how they vary day by day over the week—can help prepare the operations team to uncover hidden inefficiencies, assess passenger flow, and fine-tune schedules in the future.

The Dashboards

The dashboards that you craft to answer these business decisions should reflect a polish and simplicity that you would feel comfortable handing off to a business leader. They should be well laid-out, easy to read and navigate, with clear labels. They should present not just a single picture of how to answer the question, but provide broad insight to the business leader about their underlying question. They do not need to be interactive, but any interactivity should be clear and simple. Keep in mind that dashboards to not need to be 100% visuals either. Highlighting important numbers or groups of values might be more useful for conveying simple information at a glance.

Assessment

You will be assessed on the following for each dashboard/business decision:

Category 4 – Exemplary 3 – Proficient 2 – Developing 1 – Insufficient
Relevance to Business Question Dashboard clearly and directly answers the question; all content is purposeful. Dashboard mostly answers the question; minor gaps or tangents present. Partially answers the question; some visuals are only loosely related. Does not address the business question meaningfully.
Analytical Insight and Context Provides thoughtful analysis and useful context; insights go beyond surface-level. Provides some interpretation or context; insight is basic but present. Interpretation is minimal; context is unclear or overly simplistic. No analysis or meaningful interpretation provided.
Accuracy and Reproducibility Data is accurate and internally consistent; results could be closely replicated. Minor inaccuracies or unclear methods; mostly consistent and plausible. Not clearly reproducible; errors or inconsistencies raise concerns. Major inaccuracies or contradictions; cannot be trusted.
Design and Aesthetics Visually appealing and easy to interpret; layout, labels, and colors enhance clarity. Clear design with minor layout or labeling issues. Layout or chart choices hinder interpretation; lacks polish. Poorly designed; unclear or disorganized layout and labels.
Communication and Presentation Titles, captions, and text are clear, professional, and aid interpretation. Generally well-communicated; minor missing or unclear text elements. Explanations are vague, sparse, or unhelpful. No explanations or guiding text; confusing to the intended audience.