If you want to venture into pipelines, you can achieve this with a simple regex and loop.
If you have existing data you want to apply this to, start your pipeline with a 'search' step, then inside the loop just use an 'update record' step. And, once you update existing records and know that no other record has come during or after, then swap the search for on new event so you can update the records in near real-time. Before swapping the steps move the 'update record' step outside the loop otherwise it will be deleted when you delete the loop.
Assuming you have a plan text field in your table to put the results to, you'll switch the entry of the filed to code using '</>' and input the below:
*assumes your current field with this data is 'comments', we'll input our returned data into 'transformed comments'. Fields to update would be 'transformed comments' and then:
{% set match = a.comments | regex('(stp\w+)', true) %}
{% if match is iterable %}
{% for i in match -%}
{{ i }}
{%- if loop.last == false %}, {% endif -%}
{%- endfor %}
{% else %}
No match
{% endif %}** you may have to test for only a single 'stp...' value as we are iterating over the matched array.