Skip Clips – August 2026 Update
Hey everyone,Quick progress update on Skip Clips after a busy couple of months.
For those who don’t know, Skip Clips is an AI-powered platform that summarises long-form podcasts and videos so you can find the useful parts without listening to everything. We’ve been steadily building out the archive and features
Here's what shipped recently:
1. Redesigned homepage
Moved away from long rows of shows that made sense when we had only a few shows. The homepage now focuses on recent/featured episodes and has a clearer search bar. Also cleaned up a bunch of smaller UI issues.
It's still far from perfect but a step in right direction.
2. Hover descriptions
You can now hover over episode thumbnails to see a short summary without clicking through. Something small but meaningful.
3. Show-level performance analytics, potential new opportunity.
Added analytics on individual show pages (powered by YouTube data for now). Metrics include core audience size, 30-day views gain, hit rate, engagement rate, comment/like ratio, and growth graphs.
This is also an avenue that we are considering exploring to offer more advanced analytics on topics, keywords, shows and guests.
4. Launched the blog
Started publishing longer updates and articles here:
https://skipclips.com/blog/
Not much here at the moment, but a start is a start!
5. Growing Pains: Stronger processing pipeline
Many hours the last month were spent rebuilding the processing pipeline between the new episodes being fetched via the YouTube API, downloaded, uploaded onto the transcription server, transcribed, synced back up and finally a summary generated.
In this past for many months this was just witchcraft of many independent scripts running together with no visibility on status or troubleshooting issues, which there is many. At scale doing something right 99% of the time still leads to issues that requires effective logging and tools to track
Other similar changes were made throughout the back-end, particular to do with scaling. Shortcuts that worked for smaller numbers of items (sub 1000), now no longer work with tens of times that amount and delaying fixing it was only causing more pain!
6. Category tagging (still in progress)
This is currently one of the bigger challenges. We have over 50,000 unique tags, many of them very similar.
We’re working on organising them into proper categories so people can browse by topic instead of only by show. It’s still quite manual at the moment, but once it’s done it should also support future features like following specific topics. AI could be used, but at the moment it's more developing the category tree first and getting a better idea of the tags.
Don't blindly use automation to solve problems you don't full understand!
7. New shows added
Expanded the catalogue with several more shows, including American Thought Leaders, Into the Impossible with Brian Keating, Making Sense with Sam Harris, On with Kara Swisher, The Dr. Phil Podcast, and others.
Currently at 91 shows as of posting, 100 would make a nice number. Quality and suitability is however the aim over shear numbers.
What’s next
Get category tagging into production
Expand the number guest profiles, hopefully start experimenting with AI to see what it can do. Reliability is still very important.
Publish more blog posts. Trying to extract and write about unique useful data obtained from the platform would be nice.
Start laying the groundwork for user profiles, notifications
Traffic is still very low while we work through indexing and structure issues, but the archive and feature set keep growing. The data side feels increasingly solid.
Would love any feedback, especially around the tagging problem or what features would actually make this more useful for people.
Full Update:
https://skipclips.com/blog/august-2026-development-update
Thanks for reading.
Squawki
