I NEED HELP WITH FLOWISE CAN YOU HELP ME BUILD NESTED AGENTS (like if one agent generates a response the other agent should reprompt it to further improve this can go in a loop to get the best quality output)
Hi, I included a pdf file and a url in the Knowledge Base reader. Everything got green hooks. It just doesnt work well. It asks for name and email and that is pretty much it. Only if I ask for the url of the web site it gives a correct answer. Any idea what could be the reason? Thank you ...
Leon, As usual, it's an excellent video. I use VectorShift for some of my clients. The only real problem is the URL in the embedded code. I would like it to be truly white-label. I would also like to see a similar video for Flowise, especially the recursive web scrapping. Wayne
The landscape of automation/chatbots platforms is wider by the day...: botpress, mindstudio, make, n8n... i'm sticking for now to n8n,flowiseai and langfuse for metrics & monitoring because those can be self-hosted, super important for clients ;-)
@@leonvanzyl well mostly locally on servers inside client network ;-) but i have 2 instances on OVH public cloud (certainly not the cheapest path ... but its for historical reasons), using docker images
Hey Leon, great video series on Flowise. I'm just starting out with AI and your videos are very helpful. I am a bit stuck however and hoping you can help? Im trying to use the serp api tool to search for an image (anything) but the I'm not getting the image back in the chat window. Is there a way to do this?
Absolutely! You can pass the results of the collector to any of the available integrations.. like Airtable, Google Sheets, etc. I would probably add an LLM in that flow as well, just to check that the fields in the collector actually have values.
Hi Leon, very interesting!!! Now, how is this different from building a chat bot in Flowise? Here I See things that are already solved when it comes to directed behaviors, which in Flowise are not present or Even possible. Here I don't have to deal with vector databases, chunking or embeddigs. It resembles a classic chat bot on steroids because You can add Llm models to the flow. Why would I continue buildig in Flowise, besides the fact I can host it myself? In summary, which are the benefits of building in a kind of wild platform like Flowise when You have a platform like this that seems to have You covered in most things You need to do? Thank you very much!!! Ricardo
Hi Leon, I posted a comment yesterday and somehow it disappeard. Besides telling you how good your explanation was, I was wondering what was the advantage of using Flowise for building chatbots when this platform seems to give you many of the cumbersome task you have to manually solve in flowise,already solved, like not having to deal with splitting , embedding, vector databass, etc. Why should I keep working with flowise? I guess there must be a good reason. Your wisdom will be much appreciated!! Thank you!! Ricardo
Hi Ricardo! I know it can be really hard to decide on a platform. The knowledge base solution in VectorShift is incredible. As you said, you don't have to many build the "RAG" solution. It's all there and just works. I guess Flowise gives you a lower level of control, but most people won't care. Flowise has the upper hand when it comes to agents though. Especially now with agent teams.
Yeah. It's a wonderful time of discovery, but on the other hand it is almost an act of faith to remain loyal and stick to one solution as your main tool. And tool hopping is not a very efficient practice as it prevents you to become really proficient in any of them. Kind of sad 😢
Nice explanation, please make this type project with langchain js. How to trained model with custom data, vector store and retrieve data. I am searching video for how to trained chatbot using custom data with javascript.
@@leonvanzyl I have completed basic. But not implemented in project, please make a chatbot, trained with custom data, which store in database as embedded form. It is very helpful. Thank you 🙏
Perfect timing on this video! I was tasked with building something like this for a client. This video helps alot.
You're welcome 🤗
@@leonvanzyl quick question is it possible to add text-to-speech and speech-to-text within the chatbot block?
Clear, understandable, and very helpful.. great tutorial. Thank you!
You're welcome
I NEED HELP WITH FLOWISE CAN YOU HELP ME BUILD NESTED AGENTS (like if one agent generates a response the other agent should reprompt it to further improve this can go in a loop to get the best quality output)
Great job! as always Leon thank you, would ♥ to see it in Flowise too 🙏
Hi, I included a pdf file and a url in the Knowledge Base reader. Everything got green hooks. It just doesnt work well. It asks for name and email and that is pretty much it. Only if I ask for the url of the web site it gives a correct answer. Any idea what could be the reason? Thank you ...
Leon, As usual, it's an excellent video. I use VectorShift for some of my clients. The only real problem is the URL in the embedded code. I would like it to be truly white-label.
I would also like to see a similar video for Flowise, especially the recursive web scrapping.
Wayne
+1 for Flowise with recursive web scrapping tutorial
Thanks again Leon
hello, I already have a well-tuned model that I can use, I have the mxtral8X7B. and for the pricing remains the same, even if I use my model??
Hello Leon, this platform is not self-hosted like Flowise (free). What advantages does it offer compared to Flowise for migrating?
Leon this is very nice video , thank you
You're welcome 🤗
@@leonvanzyl Please make video in Vectorshift about User authentication with save ID and Password in airtable
Which would you choose for production if you had to only choose one? Flowise, Langflow or Vectorshift?
or Dify ???? or another one??
The landscape of automation/chatbots platforms is wider by the day...: botpress, mindstudio, make, n8n... i'm sticking for now to n8n,flowiseai and langfuse for metrics & monitoring because those can be self-hosted, super important for clients ;-)
n8n is really cool and fun to use as well! Where do you host your n8n instances by the way?
@@leonvanzyl well mostly locally on servers inside client network ;-) but i have 2 instances on OVH public cloud (certainly not the cheapest path ... but its for historical reasons), using docker images
@@christiansard5247 trop cher OVH ;)
Hey Leon, great video series on Flowise. I'm just starting out with AI and your videos are very helpful. I am a bit stuck however and hoping you can help? Im trying to use the serp api tool to search for an image (anything) but the I'm not getting the image back in the chat window. Is there a way to do this?
Hey Leon, Can we all integrate the data collector to Google sheets so that the name, email can be seen there?
Absolutely! You can pass the results of the collector to any of the available integrations.. like Airtable, Google Sheets, etc. I would probably add an LLM in that flow as well, just to check that the fields in the collector actually have values.
Hi Leon, very interesting!!!
Now, how is this different from building a chat bot in Flowise?
Here I See things that are already solved when it comes to directed behaviors, which in Flowise are not present or Even possible. Here I don't have to deal with vector databases, chunking or embeddigs. It resembles a classic chat bot on steroids because You can add Llm models to the flow.
Why would I continue buildig in Flowise, besides the fact I can host it myself?
In summary, which are the benefits of building in a kind of wild platform like Flowise when You have a platform like this that seems to have You covered in most things You need to do?
Thank you very much!!!
Ricardo
Hi Leon, I posted a comment yesterday and somehow it disappeard. Besides telling you how good your explanation was, I was wondering what was the advantage of using Flowise for building chatbots when this platform seems to give you many of the cumbersome task you have to manually solve in flowise,already solved, like not having to deal with splitting , embedding, vector databass, etc. Why should I keep working with flowise? I guess there must be a good reason. Your wisdom will be much appreciated!! Thank you!! Ricardo
Hi Ricardo!
I know it can be really hard to decide on a platform.
The knowledge base solution in VectorShift is incredible. As you said, you don't have to many build the "RAG" solution. It's all there and just works.
I guess Flowise gives you a lower level of control, but most people won't care. Flowise has the upper hand when it comes to agents though. Especially now with agent teams.
Yeah. It's a wonderful time of discovery, but on the other hand it is almost an act of faith to remain loyal and stick to one solution as your main tool. And tool hopping is not a very efficient practice as it prevents you to become really proficient in any of them. Kind of sad 😢
nice
Nice explanation, please make this type project with langchain js. How to trained model with custom data, vector store and retrieve data. I am searching video for how to trained chatbot using custom data with javascript.
Thank you.
I actually have an entire playlist on my channel dedicated to Langchain JS with RAG.
th-cam.com/video/MaSynwSIty4/w-d-xo.html
@@leonvanzyl I have completed basic. But not implemented in project, please make a chatbot, trained with custom data, which store in database as embedded form. It is very helpful. Thank you 🙏