Reports & Analytics

Reports and Analytics provide visibility into case metrics, team performance, and workflow efficiency. Aceryx offers built-in reports, natural language querying, and custom SQL-based reporting.

Built-In Reports

Aceryx includes a suite of pre-built reports accessible from the Reports menu.

Cases Summary

Aggregates case counts and statistics by week, case type, or status.

Metrics:

• Total cases created.
• Cases by status (Open, In Progress, Completed, Cancelled).
• Average case age (time from creation to completion or current state).
• Trend line showing case volume over time.

Filters:

• Date range (last week, last month, last quarter, custom).
• Case type.
• Status.

Use cases:

• Understand case volume trends.
• Identify slow-moving case types.
• Monitor backlog growth.

SLA Compliance

Tracks adherence to Service Level Agreements on human tasks.

Metrics:

• Percentage of tasks completed on-time.
• Percentage of tasks in warning (approaching deadline).
• Percentage of tasks breached (exceeded deadline).
• Escalation rate.

Filters:

• Date range.
• Case type.
• Assigned user or role.

Visualization:

• Pie chart: on-time vs. warning vs. breached.
• Trend line: compliance improving or degrading over time.

Use cases:

• Monitor team SLA performance.
• Identify bottlenecks (case types with high breach rates).
• Evaluate staffing needs.

Case Ageing

Shows how long cases remain in each workflow stage.

Metrics:

• Average duration per stage.
• Median duration per stage.
• 90th percentile duration (longest-running cases).

Breakdown:

• By case type.
• By stage (workflow step name).
• By assigned user or role.

Visualization:

• Bar chart: average time in each stage.
• Box-and-whisker: distribution of times (showing outliers).

Use cases:

• Identify bottleneck stages.
• Optimize slow stages (add resources, simplify process).
• Benchmark against historical performance.

Cases by Stage

Snapshot of case distribution across workflow stages.

Metrics:

• Count of cases in each stage.
• Percentage breakdown.
• Cases stuck (no activity in 7+ days, configurable).

Filters:

• Date range (cases created or modified in range).
• Case type.

Visualization:

• Pie chart or stacked bar chart by stage.
• Highlight “stuck” cases in red.

Use cases:

• Monitor workflow distribution.
• Spot bottlenecks visually.
• Identify cases requiring manual intervention (stuck cases).

Workload (Tasks per User)

Distributes task assignments across the team.

Metrics:

• Task count per user or role.
• Average SLA completion time per user.
• Tasks completed per user (completed count).
• Current open task count.

Breakdown:

• By user.
• By role.
• By case type.

Visualization:

• Bar chart: workload per user.
• Scatter plot: task volume vs. average completion time.

Use cases:

• Balance workload across team members.
• Identify overloaded or underutilized staff.
• Evaluate performance (task volume and SLA compliance).

Decisions (Outcome Patterns)

Analyzes patterns in task outcomes and workflow decisions.

Metrics:

• Approval rate (percentage of tasks with “Approved” outcome).
• Rejection rate.
• Request for information rate (if this outcome is used).
• Outcome distribution across case types.

Breakdown:

• By case type.
• By assignee.
• By time period.

Visualization:

• Pie chart: outcome distribution.
• Trend line: approval rate changing over time (e.g., stricter approvals in recent weeks).

Use cases:

• Detect approval bias or inconsistency.
• Monitor decision quality.
• Identify case types with high rejection rates (may need process improvement).

Natural Language Queries

Ask questions about your data in plain English. Aceryx uses an LLM to generate and execute SQL queries.

API:

POST /reports/ask
{
  "question": "How many loan applications were completed last week?"
}

Process:

1. The LLM reads your question.
2. It has access to the schema (cases table, steps table, audit log, etc.).
3. It generates a SQL query.
4. Aceryx executes the query.
5. The result is returned with a human-readable summary.

Example questions:

• “What is the average time for a loan application to reach approval?”
• “Which case type has the highest SLA breach rate?”
• “Show me cases created by Alice that are still in the approval stage.”
• “How many tasks were escalated this month?”

Limitations:

• The LLM must generate valid SQL. Complex or ambiguous questions may produce incorrect queries.
• Results are limited to 10,000 rows (configurable).
• Queries run on a read replica if one is configured, to avoid impacting production.

Custom Reports

Create and save custom reports using SQL.

Creating a custom report:

1. Click New Report.
2. Write a SQL query against the Aceryx schema (cases, case_data, steps, case_steps, tasks, audit_log, etc.).
3. Test the query.
4. Configure visualization type (table, bar, line, pie, number).
5. Save with a name and description.

Available tables:

• cases: Case metadata (id, case_number, case_type, status, created_at, updated_at, version, tenant_id).
• case_data: JSON case data (case_id, data, version).
• case_steps: Workflow step execution (case_id, step_id, type, status, activated_at, completed_at, outcome).
• audit_log: All events (id, case_id, action, actor, timestamp, details).

Example query:

SELECT
  c.case_number,
  c.case_type,
  COUNT(t.id) AS task_count,
  AVG(EXTRACT(EPOCH FROM (t.completed_at - t.activated_at)) / 3600) AS avg_hours_per_task
FROM cases c
LEFT JOIN steps s ON c.id = s.case_id
LEFT JOIN tasks t ON s.id = t.step_id
WHERE c.created_at >= NOW() - INTERVAL '30 days'
GROUP BY c.case_number, c.case_type
ORDER BY avg_hours_per_task DESC;

Materialized Views

Aceryx pre-computes materialized views for fast aggregation, enabling efficient dashboards and reports. These are separate from standard summary views.

Built-in Report Views

Three primary materialized views are provided for reporting:

• mv_report_cases: Summarizes case metrics by type, status, and date.
• mv_report_steps: Summarizes step execution metrics.
• mv_report_tasks: Summarizes task metrics including SLA compliance.

The report_view_schemas table describes the available columns in each view.

Refreshed: Periodically via background job.

Usage:

Report views support natural language querying via the API. You can also query them directly:

SELECT case_type, status, COUNT(*) as count
FROM mv_report_cases
WHERE created_date >= NOW() - INTERVAL '30 days'
GROUP BY case_type, status;

Saved Reports

Custom reports are stored in the saved_reports table with the following structure:

• original_question: The natural language question (if generated via LLM).
• query_sql: The SQL query to execute.
• visualisation: The visualization type (table, bar, line, pie, number).
• columns: Explicit column selection.
• parameters: Query parameters for filtering.
• is_published: Whether the report is available to all users.
• pinned: Whether the report is pinned to the dashboard.
• schedule: Optional cron expression for scheduled execution.
• recipients: Email recipients for scheduled reports.

Report Execution

On Demand

Click a report in the Reports menu. The system:

1. Executes the query (or fetches from materialized view if available).
2. Renders the result as a table, chart, or both.
3. Provides export options (CSV, Excel, PDF).

Scheduled

Create a scheduled report to run automatically and email results to stakeholders.

Configuration:

Report: SLA Compliance
Schedule: Every Monday at 9 AM
Recipients: manager@example.com, ops-team@example.com
Format: PDF with embedded charts

Process:

1. At the scheduled time, Aceryx executes the report query.
2. Renders the result as PDF (with charts).
3. Sends via email to configured recipients.

Retention:

Executed reports are stored for 30 days (configurable) and can be viewed in the Report History.

Dashboard

The Dashboard (main page) provides an at-a-glance overview of key metrics.

Default widgets:

• Cases by status: Pie chart of Open, In Progress, Completed, Cancelled.
• Recent cases: List of 5 most recently created/modified cases.
• SLA summary: Counts of on-track, warning, and breached tasks.
• Tasks assigned to you: Quick access to your inbox.
• Weekly case volume: Trend line of cases created per week.
• Team workload: Bar chart of tasks per team member.

Customization:

Users can customize the dashboard by:

1. Clicking Edit Dashboard.
2. Adding/removing widgets.
3. Resizing and reordering.
4. Setting filters (e.g., show only loan cases, show only my team).

Refresh:

The dashboard updates every 30 seconds by default (configurable). Use the Refresh button for immediate update.

Report Performance

For large datasets, reports may be slow. Optimize by:

1. Using materialized views when possible (pre-aggregated data).
2. Adding date filters to limit the dataset (e.g., last 30 days instead of all-time).
3. Running custom reports on a read replica (if configured).
4. Scheduling heavy reports during off-peak hours.

Monitoring:

• Dashboard shows query execution time.
• Slow queries (>5 seconds) trigger warnings.
• Logs record all report executions for audit and debugging.

Export and Integration

Export formats:

• CSV: Comma-separated values (spreadsheet-compatible).
• Excel: XLSX format with formatting and multiple sheets.
• PDF: Formatted report with charts and styling.
• JSON: Raw data for programmatic consumption.

API:

GET /reports/{report_id}/execute?format=csv&start_date=2026-03-01&end_date=2026-03-31

Email delivery:

Configure scheduled reports to email recipients automatically. Reports are attached as files (CSV, Excel, or PDF).

Webhooks:

Send report results to external systems (data warehouse, BI tools, Slack) via outbound webhooks on completion.