Task 1 commits

[abhirampal16]

For data visualization, I would rely on Looker (or a similar modern BI tool) for both the semantic layer and the final dashboards, ensuring a unified and trusted analytics environment.

The LookML semantic layer would sit directly on top of our curated dbt models, giving the Execution team a single, consistent source of truth for all calculated metrics such as gross proceeds, average open duration, and success rates. Centralizing these definitions guarantees consistency across all reports at Hiive and eliminates manual calculation errors.

For the visualization strategy, I would design a Transaction Health Dashboard focused on operational efficiency. A Sankey diagram or funnel chart would serve as the primary visualization to map transaction flow across stages—from Opened to Paid or Terminated—clearly revealing volume drop-offs and bottlenecks.

To support deeper operational insights, I would incorporate additional visuals:

• Box plots to show medians and outliers for time spent in key stages, supporting SLA monitoring
• Bar charts to break down termination reasons, enabling quick triage and prioritization
• Time-series charts to track trends over time, built using a standard date dimension for consistency

This multi-layered approach gives the team visibility into both high-level patterns and stage-level failures, allowing them to monitor performance and act quickly on operational issues.

[abhirampal16]

Execution run