How to summarize "email parser" body/content with AI?
Status: Open · Asked by BNCW Enterprises on · 0 views
My current workflow is:
- Receive an email report.
- Use Email Parser + Text Parser to extract impressions, comments, date, etc.
- Send the parsed data to Slack.
However, this often breaks because the email format changes.
Is there a way in Pabbly Connect to:
- Capture the full email content as plain text,
- Use AI (like OpenAI integration) to summarize or extract the metrics I need,
And then send that summary to Slack / next in the workflow?
If yes, what would be the best way to set this up so it’s more reliable than parsing?
Also, is it possible to have the AI compare the current email parsed content with the previous one to generate a comparison summary as well?
Hi @adityarajsingh,
My current workflow is:
- Receive an email report.
- Use Email Parser + Text Parser to extract impressions, comments, date, etc.
- Send the parsed data to Slack.
However, this often breaks because the email format changes.
Is there a way in Pabbly Connect to:
- Capture the full email content as plain text,
- Use AI (like OpenAI integration) to summarize or extract the metrics I need,
And then send that summary to Slack / next in the workflow?If yes, what would be the best way to set this up so it’s more reliable than parsing?
You can try using OpenAI to generate structured outputs of the email summary, but we won't be able to comment on its reliability as it is a large language model and may be inconsistent. However, you can set up a new workflow and test the same on your end.
youtube.com/watch?v=tB28CCtDRnE
Also, is it possible to have the AI compare the current email parsed content with the previous one to generate a comparison summary as well?
Unfortunately, this particular use case is not possible.
Thanks, trying that, does Pabbly have any plans to introduce memory module d for the AI integrations? (like in n8n)
Hello @adityarajsingh,
Thanks for trying that out.
Regarding your query, Pabbly already provides a memory-like capability within the AI Assistant by Pabbly module. Under the Create Assistant and Ask Assistant actions, you will find an option to add a Chatbot ID. By passing the same Thread ID, the assistant can recall previous interactions in that thread, effectively giving it contextual memory of the past conversation. This allows the assistant to build upon earlier chats and respond more intelligently in subsequent queries.
This is interesting, gonna try it out, thanks!
Hello @adityarajsingh,
Sure, try it out and feel free to reach out to us if you face any issues in the future.