Hey! Thank you for the comment. I will try to make content as much as possible, not sure if i can do daily since I have businesses to run, but I'll make sure I do my best!
awesome video, and I love that you did the versioning programmatically... I have a question regarding this setup though... how would you handle if a file is deleted from the watch folder because you no longer want it in the RAG db?
Hi! Thanks for the comment! The programmatic way is must for production. People should only add ai to things if code can’t do them! Regarding the automatic deletion, it should be possible.
Hi Frank, great channel! I appreciate you sharing your content with us. Would you consider creating a video or providing advice on the monthly subscriptions you recommend? Specifically, I’d be interested in learning more about API costs, self-hosted options, and local solutions. Love from Paris.
Hey! Absolutely, make sure to add all these in the google form that I have where you can register interest for the community if you are interested in joining. Otherwise you can also add this on the google form for video suggestions. Thanks for your comment :)
Hey, thank you so much for the comment! Yep, you can absolutely do this on Supabase's free plan! The free tier is pretty generous and covers everything you'd need to set this up, including storing the chat memory and vectors. Just keep in mind that if your usage scales up (like larger datasets or frequent uploads), you might eventually need to upgrade. But for most cases, the free plan should work perfectly!
Hey, thanks for the comment! That’s a great question! From my experience, companies are starting to adopt tools like n8n, especially for prototyping and integrating workflows quickly. While code is still the go-to for complex, large-scale AI projects, no-code/low-code tools like n8n are increasingly being used for tasks like automating processes, managing data pipelines, or experimenting with AI models. These frameworks make it easier for teams to collaborate and move fast without always relying on developers. So while it might not replace code entirely, it’s definitely carving out a space, especially for things like integrations and rapid deployment.
Hello again !, I’m following your tutorial on n8n using Supabase, vectors, etc. I noticed that in your video, the Session ID field in the Postgres Chat Memory node is set to “Take from previous node automatically”. However, in my setup, I don’t see that option so I choose the option available " Connected chat trigger node ". I’ve imported the template you provided, so everything else should be the same. Could this be related to a specific version of n8n, or do I need to set up something additional for the Session ID field to work as shown in your video? Right now I can´t continue and I have the following error " Error in sub-node 'Postgres Chat Memory' ENETUNREACH " Thank you!
Thanks Frank, great video as always.
Thank you, appreciate your comment :)
We need more videos like these. Plz try to upload content daily.
Hey! Thank you for the comment. I will try to make content as much as possible, not sure if i can do daily since I have businesses to run, but I'll make sure I do my best!
Thanks you for the awesome video
Hey Future Me, interesting name haha. Thanks for your comment, appreciate it lots.
Love it really, keep them coming man ❤
Thank you Will, will do! :)
Interesting and somewhat helpful too!
For the life of me i can't find the vector store tool in n8n cloud. Did it get deprecated or change name?
In dire need of a guide on how to build an n8n fully local advanced RAG system, especially dealing with codebase.
Noted on my list! I want to do a video on deploying n8n locally.
awesome video, and I love that you did the versioning programmatically... I have a question regarding this setup though... how would you handle if a file is deleted from the watch folder because you no longer want it in the RAG db?
Hi! Thanks for the comment! The programmatic way is must for production. People should only add ai to things if code can’t do them! Regarding the automatic deletion, it should be possible.
Hi Frank, great channel! I appreciate you sharing your content with us. Would you consider creating a video or providing advice on the monthly subscriptions you recommend? Specifically, I’d be interested in learning more about API costs, self-hosted options, and local solutions. Love from Paris.
Hey! Absolutely, make sure to add all these in the google form that I have where you can register interest for the community if you are interested in joining. Otherwise you can also add this on the google form for video suggestions. Thanks for your comment :)
What is your reason for managing contacts in a Doc vs a Sheet? Or are you exporting contacts to the doc from another system
No specific reason to be honest! It can be a google sheet or any other file supported by the automation. Hope my answer helps!
Amazing content, thank you so much!
Is it possible to do this on a free plan?
Hey, thank you so much for the comment! Yep, you can absolutely do this on Supabase's free plan! The free tier is pretty generous and covers everything you'd need to set this up, including storing the chat memory and vectors. Just keep in mind that if your usage scales up (like larger datasets or frequent uploads), you might eventually need to upgrade. But for most cases, the free plan should work perfectly!
Super Frank, thank you so much!
I really enjoy these videos!
would it be easier to use chatgpt assistant for the RAG agent?
Sure, it's easier! Easier but you have less control. A RAG agent in n8n gives you more control on how you can design the RAG process itself.
Im working as a data scientist. Will company ever adapt such frameworks like n8n for their work in AI or it will be only code?
Hey, thanks for the comment! That’s a great question! From my experience, companies are starting to adopt tools like n8n, especially for prototyping and integrating workflows quickly. While code is still the go-to for complex, large-scale AI projects, no-code/low-code tools like n8n are increasingly being used for tasks like automating processes, managing data pipelines, or experimenting with AI models. These frameworks make it easier for teams to collaborate and move fast without always relying on developers. So while it might not replace code entirely, it’s definitely carving out a space, especially for things like integrations and rapid deployment.
@franknillard thanks 👍
What is the piano music at the beginning please
Hii; aaa i forgot, it's definitely Chopin but it has to be a Nocturne for sure. Hope that helps haha
@@franknillard Thanks Frank
my pleasure
Hello again !, I’m following your tutorial on n8n using Supabase, vectors, etc. I noticed that in your video, the Session ID field in the Postgres Chat Memory node is set to “Take from previous node automatically”. However, in my setup, I don’t see that option so I choose the option available " Connected chat trigger node ". I’ve imported the template you provided, so everything else should be the same. Could this be related to a specific version of n8n, or do I need to set up something additional for the Session ID field to work as shown in your video? Right now I can´t continue and I have the following error " Error in sub-node 'Postgres Chat Memory' ENETUNREACH "
Thank you!
Hi! Please can you post this on the free Discord Community, so you can also post pictures. Thank you :)
Can you send the link?
🥣 FREE RESOURCES:
Link to the presentation & n8n agent for FREE:
Done, thank you for spotting that! :)