The question seems a bit broad to give concrete solutions... You should add as much detail as possible to be able to receive better help.
Can you add examples of actual log lines and keywords you have to search?
Do the keywords follow a coherent format?
Do you have to capture a fixed list of keywords or any unknown one that follow the expected format, like "WhateverUnknownException"
Do you work with multiline logs?
The keywords appear only once per log line or multiple keywords can be present in the same line?
Given the responses to these questions, maybe a simple logstash processing can be applied to extract a keyword to a dedicated field whenever it's present, add tags or any other solution.
In the simplest approaches, even an easy (but probably less efficient) kibana visualization can be built using filters aggregation that separates the required terms without any input processing.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.