Improve CI/CD Pipeline Analytics: Add relative percentages and flexible time grouping for failure ratio visualization

Summary

The current CI/CD Pipeline Analytics dashboard displays absolute counts of successful and failed jobs per day. This is good if you're interested in volume of jobs executed, but makes it difficult to understand the actual failure ratio trend over time.

Problem

1. Absolute numbers obscure trends: When viewing raw counts of successful vs. failed jobs, it's hard to understand whether the failure rate is improving or degrading, especially when total job volume fluctuates
2. Fixed daily granularity: The X-axis is locked to daily intervals, which can be too granular for identifying meaningful patterns over longer time periods
3. No relative comparison: Without percentage-based views, teams cannot easily answer questions like "What percentage of our jobs failed this week compared to last week?"

Proposal

1. Add percentage/relative view option

Allow users to toggle between:

• Absolute counts (current behavior): Show raw numbers of successful/failed jobs

• Percentages: Show success rate and failure rate as percentages of total jobs as a stack area graph that tops at 100%. Something like:

  

2. Add flexible time grouping

Allow users to select X-axis granularity:

• Per day (current behavior)
• Per week: Aggregate data into weekly buckets
• Per month: Aggregate data into monthly buckets

Expected Benefits

• Easier identification of failure rate trends regardless of job volume changes
• Better signal-to-noise ratio when analyzing longer time periods
• More actionable insights for teams trying to improve pipeline reliability

Additional Context

This enhancement would help teams answer questions like:

• "Is our failure rate trending up or down?"
• "How does this month's reliability compare to last month?"
• "Are we meeting our target of less than X% pipeline failures?"