Sprint 6: Analytics (Weeks 13–14)

Admins can manage the platform. Sprint 6 gives them insight into it. Data visualization turns raw database counts into charts that reveal hiring trends, application velocity, and job performance.

Sprint goal

The admin dashboard includes interactive charts showing: applications per job, application status breakdown, and new applications over time.

What each role ships this sprint

Backend

Add a JSON API for the analytics data (separate from the HTML routes):

GET /admin/analytics/applications-per-job
    → [{"job_title": "Backend Engineer", "count": 14}, ...]

GET /admin/analytics/status-breakdown
    → {"pending": 34, "reviewing": 12, "accepted": 5, "rejected": 21}

GET /admin/analytics/applications-over-time
    → [{"date": "2026-09-01", "count": 3}, ...]

These endpoints must be protected by require_admin and return JSON (not HTML).

Frontend

• Add a "Analytics" tab to the admin dashboard
• Fetch data from the three endpoints above using fetch() or HTMX hx-get with hx-swap="none" + a JS callback
• Render three charts using Chart.js (CDN link, no npm needed):
  • Bar chart: applications per job
  • Doughnut chart: status breakdown
  • Line chart: applications over time

QA

• Test each analytics endpoint with: no data, one data point, many data points
• Test that non-admins receive 403 from all analytics endpoints
• Do not test the chart rendering itself (that requires a browser) — test the data endpoints only

DevOps

• Add a database backup script scripts/backup_db.sh that copies jobboard.db to a dated file
• Document the backup strategy in CONTRIBUTING.md

Tech Lead

• Decide the date aggregation granularity for "applications over time" (daily? weekly?)
• Review all analytics PRs to ensure the SQL queries use indexes and will not slow down on large datasets
• Open a "performance" issue: what happens if we have 10,000 applications?

PM

• Write a "what would you want to see?" issue and gather input from the team before backend starts work
• Write the sprint retrospective: what went well, what was hard, what would you change?

Chart.js quickstart

<!-- In templates/admin/analytics.html -->
<canvas id="applications-per-job"></canvas>

<script src="https://cdn.jsdelivr.net/npm/chart.js"></script>
<script>
  fetch('/admin/analytics/applications-per-job')
    .then(r => r.json())
    .then(data => {
      new Chart(document.getElementById('applications-per-job'), {
        type: 'bar',
        data: {
          labels: data.map(d => d.job_title),
          datasets: [{ label: 'Applications', data: data.map(d => d.count) }]
        }
      });
    });
</script>

References

• Chart.js getting started
• SQLModel group by
• FastAPI JSON responses

Sprint 6 checklist

• [ ] All three analytics endpoints return correct data
• [ ] All three charts render on the admin dashboard
• [ ] Analytics endpoints return 403 for non-admins
• [ ] Tests cover empty-state, single, and many-record cases
• [ ] Sprint retrospective is posted in Canvas
• [ ] You have 2+ merged PRs
• [ ] You have given 2+ reviews