Maximize Insights from Gong Call Transcripts in Dust Folder

·Jul 03, 2025 09:29 AM·

Hello, We’ve got thousands of Gong call transcripts saved in a Dust folder. Each transcript starts with the account name and ID (we set it up this way ourselves, not with the Gong integration, so we could add some extra metadata). I’m trying to figure out the best way to pull out useful insights from these transcripts, but only for a specific account. I’ve played around with the Search & Extract tools, but the results aren’t always tied to the account I’m looking at. I also tried querying our Snowflake table, but that doesn’t really let us do any semantic searching in the transcript text. Any tips on how to reliably filter and analyze transcript docs for just one account (using the account name or ID), and then run searches or extract insights from only those docs? I’d love some advice on the best workflow for this in Dust!

3 comments

· Sorted by Oldest

Stephen (Dust)
·
·
Hi Denis Kirchen 👋 I think you could use the filter feature and add the account name as a new label in this use case! Hope this helps! https://www.youtube.com/watch?v=QwHyMpVJva4&ab_channel=Dust-Transformhowworkgetsdone.
Denis Kirchen
·
·
Stephen (Dust) Thx for your response & would be awesome if that works 🙏 I see that your native integration with Gong now adds several labels. However the account name or SF account id, are missing. What would be the best practice to auto set these labels dynamically for each transcript document inserted in the Dust knowedge base ?
Stephen (Dust)
·
·
If the labels are not synced from the Dust native integration, I think you'll have to edit the files after they've been added to Dust, adding the tag programatically, using the upsert document endpoint

Stephen (Dust)
·
·
Hi Denis Kirchen 👋 I think you could use the filter feature and add the account name as a new label in this use case! Hope this helps! https://www.youtube.com/watch?v=QwHyMpVJva4&ab_channel=Dust-Transformhowworkgetsdone.
Denis Kirchen
·
·
Stephen (Dust) Thx for your response & would be awesome if that works 🙏 I see that your native integration with Gong now adds several labels. However the account name or SF account id, are missing. What would be the best practice to auto set these labels dynamically for each transcript document inserted in the Dust knowedge base ?
Stephen (Dust)
·
·
If the labels are not synced from the Dust native integration, I think you'll have to edit the files after they've been added to Dust, adding the tag programatically, using the upsert document endpoint