Try out Flowise Cloud for free for 14 days. Remember to use my affiliate link to skip the waiting list and support my channel 🙏. flowiseai.com/auth/signup?referralCode=LEONVZ
Another great tutorial Leon :) I have installed Flowise and n8n on Render as per your video tutorial. So now you got me thinking if better to just use your presented alternative.
Thank you for the feedback! Like I mention in the video, it might be worth it if you've reached the limits of the starter plan on Render or Railway, or if those exclusive features are important to you 😊.
Leon, I noticed that out of the blue custom tools stopped being called by the tool agent properly, can’t figure out a solution after having same setup work flawlessly for months… do you think they are deactivated some features in open source to push people to cloud! I feel if I try my flow on cloud it will work again
Hey there, I don't think they would do that. I use custom tools on my flows as well. Will have a look as well Try adjusting your system prompt. Add some instruction to use the tool whenever the user asks for XYZ. I also find that the Tool Agent provides better responses than the QA Retrieval Chain, so you might want to try that.
@@leonvanzyl please tell me if you figure out something I am already using tool agent and with a couple custom tools and get requests all stopped working after last time I updated … on Reddit I saw others got same issue maybe a bug or the update required some modifications I am not sure what exactly
Leon, this looks great for chatflow oriented users. What about those of us who are focused on agentflows? Updating is only a few keystrokes, we already use our recursive text splitters to save to Pinecone and Langsmith is available within Flowise for debugging. So if you are not deploying your work externally, what is the advantage? The only thing I see is maybe an increase in messaging limits, possibly.
This is really intended for someone who does not want to struggle with hosting the service themselves and simply want a frictionless path to getting Flowise up and running.
Thanks for another informative video Leon. I have a quick question how do the predictions get calculated when they are being used by agents. What I mean to say is for example we have multiple agents and if the output of one agent gets passed as an input to another agent, does it gets calculated as one prediction? Or no matter how many agents are there only the final output is calculated as one prediction, I hope you’re getting what I mean to say.
Hi Leon, I hope you’re doing well, once again thanks for all the informative videos. I’m planning to sign up for the Flowise Cloud, but I do have few questions that are listed below: 1. Is there any limitation on how many Chatflows or Agentflows or Document Stores we can create on the Starter $35 plan? 2. Do we still need to make use of vector store such as Pinecone to store embeddings? 3. Do we still need to make use of PostgreSQL DB service such as Supabase for Record Manager? 4. Does it support using a custom domain and white label the dashboard with our own logo and branding? If you could please make a detailed video addressing these topics of Flowise Cloud should be really helpful. Many Thanks!
wow very informative video leon! question, is the chatbot`s answers also updated whenever the reference data source e.g.(excel, pdf, website) has new data?
Thank you for the feedback! The knowledge base / document stores are not automatically refreshed. Flowise provides an API that you can trigger to do so (or you'll have to Upsert manually). I'll create a tutorial on how to do this 👍
That reminds me. I've been meaning to create a tutorial on the new APIs provided by Flowise, which includes updating the document stores. Will hopeful answer your question soon.
LOL! Don't worry, we'll use the local / self hosted version in future tutorials! People were curious about the cloud service, so I wanted to cover it 👍.
@@leonvanzyl I went to the site to ask some basic questions and they don't have an assistant on their site that can answer anything. I mean if that's one of their prime uses you think they would be their on customer. Don't trust companies that don't believe in the problem they are selling
That's literally the only differences, a pretty worthless evaluator. Being able to style and customise Flowise embed to at least be able to style it that same as the test chat inside Flowise UI would be worth 35 a month, to be able to see seq agents and agent flows respond with the same styling, formatting etc as the test chat would be superb. Other than that currently using embed with Seq agents and Agent flows is trash and confusing the way the responses are presented. And still over API the response from agents, supervisor is delivered in one huge text block. They should spend more time for us to be able to customise in the way you can with Bot press for instance, rather than really not all that useful evaluators of their cloud version or god willing the Open source version! Flowise is so close to perfection!
I think the main selling point will actually be the ability to create teams and allow collaboration with team members. That and the fact that you don't have to worry about infrastructure yourself. The evaluation features are just the cherry on top in my opinion.
Let’s also remember that for companies, this price can be considered as free, so that’s fine 😊 I appreciate on my side to be able to create personal projects for nothing with the local version, with the same features as the pro version (this is almost never the case for this type of products !)
Try out Flowise Cloud for free for 14 days.
Remember to use my affiliate link to skip the waiting list and support my channel 🙏.
flowiseai.com/auth/signup?referralCode=LEONVZ
Another great tutorial Leon :) I have installed Flowise and n8n on Render as per your video tutorial. So now you got me thinking if better to just use your presented alternative.
Thank you for the feedback!
Like I mention in the video, it might be worth it if you've reached the limits of the starter plan on Render or Railway, or if those exclusive features are important to you 😊.
@@leonvanzyl thank you and looking forward to your next awesome tutorials :)
Hi leon, can you also host a local model within the flowise cloud platform? That would be 😎
you are the best youtuber ever among all
Thank you 😊.
Just trying to be helpful and provide value to you guys.
Leon, I noticed that out of the blue custom tools stopped being called by the tool agent properly, can’t figure out a solution after having same setup work flawlessly for months… do you think they are deactivated some features in open source to push people to cloud! I feel if I try my flow on cloud it will work again
Hey there, I don't think they would do that.
I use custom tools on my flows as well. Will have a look as well
Try adjusting your system prompt. Add some instruction to use the tool whenever the user asks for XYZ.
I also find that the Tool Agent provides better responses than the QA Retrieval Chain, so you might want to try that.
@@leonvanzyl please tell me if you figure out something I am already using tool agent and with a couple custom tools and get requests all stopped working after last time I updated … on Reddit I saw others got same issue maybe a bug or the update required some modifications I am not sure what exactly
Leon, this looks great for chatflow oriented users. What about those of us who are focused on agentflows? Updating is only a few keystrokes, we already use our recursive text splitters to save to Pinecone and Langsmith is available within Flowise for debugging. So if you are not deploying your work externally, what is the advantage? The only thing I see is maybe an increase in messaging limits, possibly.
This is really intended for someone who does not want to struggle with hosting the service themselves and simply want a frictionless path to getting Flowise up and running.
Thanks for another informative video Leon. I have a quick question how do the predictions get calculated when they are being used by agents. What I mean to say is for example we have multiple agents and if the output of one agent gets passed as an input to another agent, does it gets calculated as one prediction? Or no matter how many agents are there only the final output is calculated as one prediction, I hope you’re getting what I mean to say.
That's actually a good question!
I'm guessing that the analysis would (or should) return the totals of the different agents.
Hi Leon,
I hope you’re doing well, once again thanks for all the informative videos. I’m planning to sign up for the Flowise Cloud, but I do have few questions that are listed below:
1. Is there any limitation on how many Chatflows or Agentflows or Document Stores we can create on the Starter $35 plan?
2. Do we still need to make use of vector store such as Pinecone to store embeddings?
3. Do we still need to make use of PostgreSQL DB service such as Supabase for Record Manager?
4. Does it support using a custom domain and white label the dashboard with our own logo and branding?
If you could please make a detailed video addressing these topics of Flowise Cloud should be really helpful.
Many Thanks!
flowise vs n8n, which one you recommend?
Both 😎.
n8n for automations and integration with external systems, and Flowise for some of the more advanced AI features.
Need video about this topic.
Leon, looks great tutorial. if possible make an video(chatbot to human handoff) like "include human agents from Microsoft teams into flowise chatbot".
Will see what I can do. Awesome idea for a tutorial.
This with the files, does it mean, there ist no need of vector stores like pinecone or qudrant anymore? That would be great!
wow very informative video leon! question, is the chatbot`s answers also updated whenever the reference data source e.g.(excel, pdf, website) has new data?
Thank you for the feedback!
The knowledge base / document stores are not automatically refreshed.
Flowise provides an API that you can trigger to do so (or you'll have to Upsert manually).
I'll create a tutorial on how to do this 👍
@@leonvanzyl yeah2 i think its the RAG method. im grateful with your video. thanks 🌟
Thank you for the tutorial. Can Flowise and n8n be installed on the same VPS? I know n8n takes all the server, dealing with traeffic stuff.
Yes, I'm doing it. You can even do it on different ports or even dockers.
😎
If i have my own cloud server hosted online, can i host flowise there and reap the same benifits for free/what i already pay for cloud hosting?
For sure. Flowise simply runs on a node server or within a Docker container.
im hosting it on k8s
How do you upsert a document to the document store through the API. It wasn’t clear when i tried. I ended up building a flow the old way.
That reminds me. I've been meaning to create a tutorial on the new APIs provided by Flowise, which includes updating the document stores.
Will hopeful answer your question soon.
@@leonvanzyl amazing Leon, I'm sure many users would appreciate a video on the new APIs
Any quick pointers? I feel like I'm close but couldn't get it work.
Dont know why, rag vector sometimes give inacurate and not relevant answer, still waiting best flow
Most likely due to the quality of your dataset or the way it has been chunked. This can have a massive impact on quality.
Here we go... the slippery slope to $$$.
LOL! Don't worry, we'll use the local / self hosted version in future tutorials!
People were curious about the cloud service, so I wanted to cover it 👍.
They don't even use it on their own site...
What do you mean?
@@leonvanzyl I went to the site to ask some basic questions and they don't have an assistant on their site that can answer anything. I mean if that's one of their prime uses you think they would be their on customer. Don't trust companies that don't believe in the problem they are selling
That's literally the only differences, a pretty worthless evaluator. Being able to style and customise Flowise embed to at least be able to style it that same as the test chat inside Flowise UI would be worth 35 a month, to be able to see seq agents and agent flows respond with the same styling, formatting etc as the test chat would be superb. Other than that currently using embed with Seq agents and Agent flows is trash and confusing the way the responses are presented. And still over API the response from agents, supervisor is delivered in one huge text block. They should spend more time for us to be able to customise in the way you can with Bot press for instance, rather than really not all that useful evaluators of their cloud version or god willing the Open source version! Flowise is so close to perfection!
I think the main selling point will actually be the ability to create teams and allow collaboration with team members.
That and the fact that you don't have to worry about infrastructure yourself.
The evaluation features are just the cherry on top in my opinion.
Let’s also remember that for companies, this price can be considered as free, so that’s fine 😊
I appreciate on my side to be able to create personal projects for nothing with the local version, with the same features as the pro version (this is almost never the case for this type of products !)