Видео

@SoloJetMan That means a lot! Thanks so much :)

Yeah theoretically n8n enables us to do almost everything voiceflow offers and much more. It may just require more number of nodes/steps and other services. Voiceflow is a managed service that enables you to create important and common agents easily and faster. If your use case and budget aligns with voiceflow, you should consider it in favor of avoiding headache to manage your own server and a number of services.

@FuturMinds I know n8n and I host my stuff. No rush either. So my question is if I I know n8n would I really need voiceflow?

I guess not but I was told otherwise

@KarthickKalidoss Can you elaborate the use case? If you want to use database for better responses, you can use SQL Agent (if you have sql db)

@@FuturMinds Lets say i have a db of appointment and timings. And, if someone asks if an appointment is available in 4-430pm slot. I should be able to read from db and tell them a slot is availble and collect details and book his details against the slot.

@HarshSharma-vm3nm I'm glad you liked it. Thanks for the suggestion :)

Glad I could help!

@ramp2011 Yes that's what OpenAI node is doing behind the scenes for transcribing the audio. We need an agent here instead of a simple openAI API call because we want to use additional tools and chat memory to prepare the final response.

@@FuturMinds Thank you so much for your response. That makes sense

@takimdigital3421 Added in the description.

@ you are the best thank you

@SoloJetMan You could create another workflow (or update existing one to route to different flow) that would either create embeddings to vector store instantly, or you could write the user inputs somewhere (e.g google sheet) and have a cron running that checks for new entries every few min or hours and creates embedding out of them. Hope this answers your question.

Glad it was helpful!

Awesome, let me know how it goes!

Glad it was helpful!

@scottzeta3067 The Assistant decides whether to call a function based on how well the user's natural language input matches the function's description, parameters, or purpose. If the input aligns closely with a specific function's description, it triggers that function. Otherwise, it handles the query generically without calling a function.

I'm glad you found it useful!

Thanks @XXAI-1

Glad you found it useful :)

Glad it was helpful!

@SoloJetMan Flowise is a pretty powerful tool and one of my favorite. It enables us to avoid writing langchain code to build poc while finding market fit and good for prototyping. But if we talk about building long term production applications we start to see the limitations. Production app is supposedly used by good number of customers and you need to keep building new features to meet the market needs. While doing that you want a really good control over your application in terms of manageability and scalability. You also want proper logging for easier debugging, metrics, trends, stats and so on. These are a few things that makes it not so suitable option for big production applications.

Thanks, you too!

Thanks for the suggestion. Will try to cover these in the future videos :)

Awesome to hear you liked it!

@dataswamibot Maybe because of the following reason: 1. You can not converted your instagram account to business account. 2. Your token doesn't have the required permissions to access this instagram page. 3. You have not linked your instagram page to your corresponding facebook page.

Awesome, glad you found it helpful!

Glad it helped

@DrBrianKeating You can check which tier you fall in for OpenAI here: platform.openai.com/docs/guides/rate-limits?context=tier-one#reduce-the-max_tokens-to-match-the-size-of-your-completions Let's say you are tier1 and use gpt-3.5-turbo, then you can make 3,500 requests per minute and 200,000 of token combined over all the requests. If the video you are analyzing has really large number of comments, then you may see the token limit errors. To get around this, you can add a sleep of say, 2 seconds in the get_comments_summaries function inside for loop. So the updated code would look like this: #import time lib import time # Function to get summaries from OpenAI def get_comments_summaries(batches): summaries = [] for batch in batches: response = client.chat.completions.create( model="gpt-3.5-turbo", messages=[ {"role": "system", "content": "Summarize the following comments while keeping the detailed context."}, {"role": "user", "content": " ".join(batch)} ] ) print(response) summaries.append(response.choices[0].message.content) time.sleep(2) #Sleep for 2 seconds return summaries

Thanks! I'm glad you liked it!

Awesome, I'm happy it was useful!

Great catch, I'll make sure to add that to github documentation.

Awesome, glad you liked the video!

Awesome, thanks!

Not possible. n8n requires long running VM with some memory and space.

@@FuturMinds HEY THANKS FOR REPLY, thanks for tutorial as well , completed my n8n installation but when i suspend my machine all things have gone also assigned with new external ip

Yes since the n8n data is stored on that same VM you terminated, it is lost. If you want your data to be independent of VM, you'll need to configure external DB.

​@@FuturMinds hello , i am able to run n8n over google cloud through your video but i got a problem like no one having, any workflow get stuck at executing and when i stopped it got executed successfully, even in chat model i have to stop after every asking question to able to run properly,. Need help please

Ok some how i managed to solve it thanks

I'm glad you found it helpful!

I'm glad you found it helpful!