While waiting around at different points over the last free weeks I have been trying to to pull together better content on this website including bringing the show notes up to date and adding in the full transcitpions of the interviews. This I hope will help with SEO and what every AI indexing stuff goes on and allow George (www.nigelcreaser.com/george) my AI agent to provide even more value.
This has required quite a few processing steps especially as I do pay for the all singing all dancing features of Riverside (affiliate link: www.nigelcreaser.com/riverside).
As these steps were quite manual I decided to see if I can speed it up.
1. Freshly downloading the audio files from the rss feed to allow transcription. Google Gemini provided some code and istructions on how to do this in the form of an chrome extension that reads an open rss feed in the browser and downloads all of the files.
2. Run each file through https://freepodcasttranscription.com a bit if a pain as each needs to be done one by one, not fou d a free slick batch alternative.
3. As the download format for text are not great I downloaded as .pdf. I then needed to create a text version, so I turned to Gemini again to give me some code and it created a .bat file for me to run the co version of multiple files through command line tool called pdftotext.
4. Finally to get the text into posts… hello again Gemini, fix my email to posts set up on my wordpress site and and a python script created to go through each .txt in a folder and create and send and email to the blog.
I am part through the update but know i have saved several hours of effort already I have some tidying up to do of each transcriptiom and creation of a decent summary (will lean on Gemini again), move the transcripts to old posts, get images and guest links etc on any new posts.
Each week I seem to be embracing the possibilities and use cases for AI.
I also suspect many of my techie eagues who have already fully embraced will be shouting “Duh! Nige.”
All the prompts, code and links to tools are here:
