Hey everyone, I’m building a visual dashboard with Lovable connected to Supabase. The main thing I’m struggling with is making it work properly for multiple clients. Basically, this is for a marketing agency with around 10–15 clients. Ideally, I’d like to have just one Lovable site where each client can log in (magic link, login, whatever) and only see their own metrics. The data comes from two sheets (one for daily metrics and one for launch metrics). Automations to update the Supabase tables are already done, and Supabase is connected to Lovable. A few things I’m not sure about: - For auth, should I use magic links or some type of manual login? - Should I set up the RLS policies myself, or can I let Lovable handle that? - What’s the best way for Lovable to query specific data from Supabase—via a prompt/LLM setup, or by telling it directly? Has anyone done something similar or has tips on the best setup?

When I try to follow Nate his course, the first workflow of pinecone vector store if I try to run it I come through an error The error message: Vector dimension 1536 does not match the dimension of the index 512 means that the embeddings you are trying to insert into your Pinecone index have a vector size (dimension) of 1536, but your Pinecone index basvida is configured to accept vectors of dimension 512. Actionable Solution:Check which embedding model you are using in your workflow. For example, OpenAI's text-embedding-ada-002 model produces 1536-dimensional vectors. If your Pinecone index is set to 512 dimensions, you must either: - Change your embedding model to one that outputs 512-dimensional vectors, or - Recreate your Pinecone index basvida with a dimension of 1536 to match your embedding output. Even tho I followed all Nate his steps correctly and I choose the right language model it still gives me the error anybody an idea on how I can solve this?

Hey fellow automators 👋 Today stripe closed my account saying your business is high risk ... even after I appealed. Anyone faced this issue before, if so how did you fix it?

My current task is to link a dropbox folder to a workflow, invoices and receipts, that need to be processed (ocr) and then manipulated into a form that can generate reports of stores, dates, items, costs. etc etc, the usual. My first thought, was that I could use the OCR from Nates InvoiceAgent video, and then submit all the data to Pinecone Vector Store, so that I could ask an agent about the receipts. I modified the "Nike Sales agent" that used pinecone. But when I run my query i get no results. Does anyone have any ideas, or better suggestions?

Does anyone knows how to upload binary file to monday.com subitem column? everything in the monday column match perfectly just the file is not able to upload. really needed help on this. 🫡