I used the OCR package/library in regular code/script Goal was to capture screen and transfer info to ChatGPT for question answers Limited in math and graphs etc. looking forward to learn n8n and agentic usage
have you tried using gemini-1.5-flash, it parses (OCR) data and you can have all parsing rules, output structure etc.. specified in one Basic LLM chain node
BIG W bro. Now the only thing we are waiting for is how to CREATE an Invoice XD. I think that is the biggest challenge tbh. Can we use the Agent to draft an invoice and PDF it with Goutenberg??
why didn't you set manually the ID to avoid the third telegram node (get:file)? It works more elegantly for referencing back nodes. So you'd have two Set nodes
That’s a great suggestion I appreciate the feedback. I guess my thought process was to show more transparency of what’s going on, especially since it doesn’t cost us anything to re-use the telegram node. I appreciate your thoughts!
Is that workflow still working for you? My telegram from photo files output looks very different and I get an Error "Bad Request: wrong file_id or the file is temporarily unavailable"
Perfect, thank you, by the way i'm looking for video how to give ai agent pull my cellphone contacts and put it into my client's database. Or just give it straight access to my cellphone contacts or google account. Is there any video about it? Thank you, you're the man 👍
I used the OCR package/library in regular code/script
Goal was to capture screen and transfer info to ChatGPT for question answers
Limited in math and graphs etc. looking forward to learn n8n and agentic usage
Thanks !!! It has been very useful. Please keep it up.
have you tried using gemini-1.5-flash, it parses (OCR) data and you can have all parsing rules, output structure etc.. specified in one Basic LLM chain node
Awesome !!! insight share Thx...
BIG W bro. Now the only thing we are waiting for is how to CREATE an Invoice XD. I think that is the biggest challenge tbh. Can we use the Agent to draft an invoice and PDF it with Goutenberg??
How would you connect to an existing personal assistant to this flow?
Cool. What about n8n security when connect to backend database? Can we apply row level security. Is n8n production ready
why didn't you set manually the ID to avoid the third telegram node (get:file)? It works more elegantly for referencing back nodes. So you'd have two Set nodes
That’s a great suggestion I appreciate the feedback. I guess my thought process was to show more transparency of what’s going on, especially since it doesn’t cost us anything to re-use the telegram node. I appreciate your thoughts!
Is that workflow still working for you? My telegram from photo files output looks very different and I get an Error "Bad Request: wrong file_id or the file is temporarily unavailable"
Perfect, thank you, by the way i'm looking for video how to give ai agent pull my cellphone contacts and put it into my client's database. Or just give it straight access to my cellphone contacts or google account. Is there any video about it? Thank you, you're the man 👍
im waiting for an agent that cant treat complex tables and deal with Chinese hand written text :D