Help building template

Grafana Elasticsearch
1

Hello,

Using python I ingest system metrics into ES (example below).

Now, I’d like to use the templating to display only the needed information, let’s say the “user” and “system” values.

{
  "_index": "system-2018-02-26",
  "_type": "system",
  "_id": "QS9l0WEBwt1FT8MlvRO6",
  "_version": 1,
  "_score": null,
  "_source": {
    "user": 9,
    "nice": 22.7,
    "system": 3,
    "idle": 64.5,
    "iowait": 0,
    "irq": 0.4,
    "softirq": 0.4,
    "steal": 0,
    "guest": 0,
    "guest_nice": 0,
    "timestamp": "2018-02-26T09:16:00.782Z",
    "tag": "cpu_percent",
    "hostname": "magnesium",
    "cpu": "core0"
  },
  "fields": {
    "timestamp": [
      "2018-02-26T09:16:00.782Z"
    ]
  },
  "highlight": {
    "tag": [
      "@kibana-highlighted-field@cpu_perc@/kibana-highlighted-field@"
    ]
  },
  "sort": [
    1519636560782
  ]
}

Do you have an advice?
maybe the ingested format does not fit the requirements, if so, what format should I use?