EventSourcingDB Analysis Services

The answers are already in your events. Getting them out usually takes a project.

Events record what happened in the language of the business – the best raw material for analysis there is. Yet the usual path runs through an export, an ETL job, a warehouse, and a data model that has to follow every change. Analysis Services skips all of that and reads the stream directly.

• The schema is inferred, not declared. Fields, types, and cardinalities come from the events themselves, including the old shapes that never go away.
• Suggestions follow from the data. The catalog of analyses is derived by rules from the inferred schema, so every suggestion can actually be answered.
• Dashboards are files. A board is a JSON file that can be reviewed, versioned, and deployed like code. The editor is one way of writing it.

{
  "version": 1,
  "title": "The library at a glance",
  "tiles": [
    {
      "id": "loans-per-month",
      "title": "Books borrowed per month",
      "chart": "line",
      "layout": { "x": 0, "y": 0, "w": 8, "h": 4 },
      "series": [
        {
          "query": {
            "from": "io.eventsourcingdb.library.book-borrowed",
            "bucket": "month",
            "measures": [{ "fn": "count" }]
          }
        }
      ]
    }
  ]
}

A board with a single tile, which draws the loans of every month, live. The editor writes the same format.