JSON-Generating Data Flow Process

  1. Every visitor and/or agent action triggers an event.
  2. Events are stored in HDFS (Hadoop File System).
  3. Job processes the events (in the Hadoop environment) and generates a grouping of all the events related to each session once they are considered "closed".
  4. A data access job then runs on the above grouped data in 1 hour intervals and generates the JSON files.

Flow

Data access is not a default function of Conversational Cloud; it must be configured separately for each account. The four types of data supported and configurable in each account are as follows:

  • Agent Activity
  • Web Session
  • Engagement
  • Survey

The Data Access API retrieves historical data. The SLA for data retrieval can be up to 6 hours from the time the data is considered "closed" (for 99.5% of the data).

This API supports the HTTPS GET protocol (data retrieval), and all retrieved data is returned in JSON format based on the AVRO schema.