Execution Heatmap

Visualise workflow execution density over time — spot bottlenecks and peak-load windows at a glance.

Overview

The Execution Heatmap renders a calendar-style grid where each cell represents a time window (hour or day depending on the selected range). Cell colour intensity reflects how many workflow executions occurred in that window — light cells are quiet periods, dark cells are busy periods.

This is the fastest way to answer questions like "when do our pipelines run most often?" or "which days have unexpectedly high failure rates?".

Reading the Heatmap

Colour scale — ranges from near-white (zero or very few executions) through amber to deep red (maximum density in the selected range).
Hover tooltip — hover any cell to see the exact execution count, success rate, and average duration for that window.
Row = workflow — each row represents a single named workflow. You can filter to show only specific workflows.
Column = time window — columns represent hours (in 24h view) or days (in 30-day / 90-day view).

Spot Unexpected Load

A single dark cell surrounded by light cells usually means a webhook burst or a scheduled batch job ran at an unusual time. Click the cell to jump to the Monitoring log filtered to that exact time window.

Filters and Ranges

Time range — 24 hours, 7 days, 30 days, 90 days.
Workflow filter — show all workflows or select specific ones from the dropdown.
Metric — switch between execution count, error rate, or average duration as the colour value.

Use Cases

Capacity planning — identify peak hours and ensure external API rate limits are not exceeded during those windows.
Incident investigation — correlate a spike in errors with a specific deployment or external event.
Scheduler validation — confirm that cron-scheduled workflows are running at exactly the times you configured in Task Scheduler.

AI Assistant

Execution Heatmap ships with a built-in AI Assistant. Click the vertical AI Assistant tab on the right edge of the screen (or the Sparkles toggle) to open a chat panel that reads the current heatmap's per-node metrics — frequency, average duration, error rate, and data volume. Because this is a dashboard/overview screen rather than a record-editing table, the assistant is scoped to dashboard-style capabilities only — it narrates and explains the numbers already shown, it never edits workflow nodes:

Executive summary — a plain-language readout of the workflow's performance profile, calling out the busiest, slowest, and most error-prone nodes by name.
Root cause & anomaly detection — identifies which specific node is the bottleneck (highest duration or error rate) and explains why it stands out relative to the others.
Recommendations — suggests concrete optimizations (parallelizing a slow node, adding retry logic to a high-error node, batching a high-volume node), referencing the actual node names and numbers.

Every reply can be saved with the Save as Report button, which stores it in My Reports under a title derived from your question and the current month/year. The panel respects the Writer Engine selected in Settings → Agentic (Local CLI, Hosted API, Local LLM, or Ollama) — Ollama and other local/in-browser engines run entirely on your own machine and never proxy through the server.

Frequently Asked Questions

What does the colour of a heatmap cell actually represent?

Colour intensity encodes the selected metric for that time window — by default, execution count. Near-white cells mean zero or very few runs, amber means moderate activity, and deep red means the highest density in the currently selected range. Switching the Metric filter to error rate or average duration re-colours the entire grid using that value instead, so the same grid can answer different questions without changing the layout.

Why do the columns change from hours to days when I switch the time range?

The column granularity automatically adapts to the selected range so the grid stays readable: the 24-hour view buckets executions by hour, while the 7-day, 30-day, and 90-day views bucket by day. This keeps the number of columns manageable and makes multi-day density trends easier to spot than a very long single-hour timeline would.

Can I see the heatmap for just one workflow instead of all of them?

Yes. Use the Workflow filter dropdown to narrow the grid to one or more specific named workflows. Each row in the heatmap always corresponds to a single workflow, so filtering simply reduces the number of rows shown — it does not change how density is calculated per cell.

How do I investigate a dark cell that looks like unexpected load?

Hover the cell first to check its exact execution count, success rate, and average duration in the tooltip. If it still looks anomalous, click the cell to jump directly into Monitoring with the log pre-filtered to that exact time window — this is the fastest path from "something spiked" to seeing the actual execution records that caused it.