CloudWatch is really powerful. As well as just searching your logs you can quickly aggregate and visualise data from your logs.
In this example we use a regex make a capture group for loglevel based on lines such as 2022-10-26T16:00:00.000Z ERROR something bad happened! then count how many times the logs contained a WARN or ERROR in the log file in a 30 minute interval. This query can be used in conjunction with the Visualization tab in CloudWatch (select Line from the dropdown).
With a little editing based on your use case, this query wiil help build an overview of your error rate
fields @timestamp, @logLevel
| parse @message /^\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}.\d+Z\s(?<@logLevel>\w+).*$/
| filter ispresent(@logLevel)
| parse @logLevel "ERROR" as errors
| parse @logLevel "WARN" as warnings
| stats count(errors) as TotalErrors,
count(warnings) as TotalWarnings
by bin(30m)