This video was a lot of effort to create. Subscribe to my channel to be notified of more FlowiseAI videos ⭐. Let me know what other videos I should create - whether it's Flowise related or not.
And as for Flowise, I would love to see a video on monetizing assistants and agents, e.g., using integration with Stripe and payment for tokens, flat fee or subscriptions.
I just came across your channel a few hours ago and I want to say a very big thank you to you. This is the first UA-cam channel in a while that is providing its users with practical knowledge and not this junk clickbait channels that are littered all over UA-cam adding zero value. That said Please make a detailed Flowise video showcasing how to create, store, fetch, and delete user data in Supabase based on each user's input on Telegram or WhatsApp. The video training that I am suggesting will get you a lot of viewership and subscribers since AI workers and agents will eventually render SaaS apps obsolete and when that time arrives users who don't want to communicate with apps via voice still need a UI by then to interact with the ai app and that UI will be either WhatsApp or Telegram.
Leon hi! When you do not have a long video, I start to worry that you are tired of telling us amazing chips, but when a new video is released, and even more so a tutorial becomes clear what a colossal work has been done for us that we can create and be happy. I and my family are grateful for your efforts and if blogodorya them I can earn my first money, I will definitely be able to buy you a coffee machine)))) thank you!
What immense value! I have been following you since the first flowise video and you have helped me so much. I run my own AI Agency now creating AI solutions and I still come to your videos every now and then to learn something new. Keep up the great work!
Wow Leon, just when you think your videos cannot get any better, Boom!! you shock the community again with amazingly easy to follow knowledge, great examples and real-life scenarios. I know we will meet one day because I want to shake the hand of someone who has inspired me to build great things. Thank you Leon! ⭐⭐⭐⭐⭐
Thank you so much, Leon, for your outstanding effort. Your wisdom has been an incredible benefit to me, and I feel like the greatest recipient of it. Thank You again.
Leon, your videos are an absolute game-changer! Your way of breaking down complex AI concepts makes it so much easier to understand and keeps me motivated on my learning journey. While I’m not quite there yet, your content makes me believe it’s possible to start building some amazing agents and web apps soon. Thank you for the consistent value you bring!
Leon, I told that person who previously said Flowise isn't production-ready due to logging, monitoring, and error handling issues. Now, he's changed his stance, but still mentions some concerns that I don't fully understand. What should I tell him, Leon? You're my expert in this, and I'm fully supporting Flowise! He said: 'I didn't know Flowise integrates with LangSmith, which addresses a major limitation. LangSmith helps with logging and monitoring, but not error handling. It's still a drawback to rely on a specific tool (LangSmith) and not have easy customization options. However, these downsides don't mean Flowise isn't worth using. I think it's a phenomenal platform!'
LOL, don't you just love debating people in the comment section 😂. I'm not sure what he means by error logging then. 1) Flowise logs any technical errors in the console which you can view in the logs if you're using something like Render. Flowise cloud has a dedicated menu options for these logs as well - making it super simple to view any errors. 2) Langsmith (and the other platforms that FW can integrate with) do offer error logging as well if the Langchain Runnables failed for any reason (including issues with tools or LLM providers).
You are doing a grate job !!! Thank you!!! I hope to see flows with API loaders and agents integrations) that would be very usefull) for example integrations between different platforms through their REST APIs) Thank you ones again!
Leon, thank you for your prompt response! I'm eager to learn from your tutorial on Pydantic AI. Could you also provide insights on why Pydantic AI stands out and how it compares to no-code tools like Flowise in terms of production deployment? Your expertise in showcasing top-tier AI automation and agent products is truly impressive, and I'm excited to gain knowledge about production-ready AI applications and AI Agents. Even if transitioning from pure no-code to minimum code, I'm interested in learning how to achieve what no-code tools can't achieve with minimal coding from you!
I actually have several videos on integrating FW with Make 😊. Since they've been dodging paying creators affiliate commission, I moved all videos over to n8n.
Hi Leon. A very happy new year. No one makes FlowiseAI contents better than you. I think we will need your help with a cheat sheet of nodes based on specific use cases. There are so many nodes available. Sometimes, it becomes overwhelming. I think we will start to see community contribution towards node documentation pages as well.
This video covers all the nodes you need to know 👍. Seriously, I know there are many nodes, but with the nodes in this video, you can achieve anything.
Awesome work! Already subscribed, but shared it among friends who are also busy with their Ai projects. Hopefully more ppl will share this. Amazing work!
Flowise is definitely a great tool for building complex scenarios. Unfortunately, it's often hard to understand everything you can do and all the capabilities offered by every component, simply because of the lack of documentation. For example, I was not able to use or even understand how to set up the additional parameters for the chat prompt template, and I have many other examples like this. I hope the documentation will improve to demonstrate the quality of the product and allow everyone to build advanced flows.
Thank you! It's a great way to kick off the new year. I usually code simple Python apps on my own, but I use Flowise for fast prototyping. It's a very useful tool! :)
Thank U So much Leon 🦁You are my favorite AI UA-camr! ...It would be great how we can interact with databases, a Classic CRUD for example. Greetings from México...
Thanks for this great work! I’d like to ask if you could create a video tutorial on using Firecrawl’s API to scrape a website with more than 200 tabs and subpages. To update a chatbot knowledge base every 24 hours. If you know of any other method that might be better suited for this scale and frequency, we’d love to hear about that as well. The key for me is ensuring the knowledge base stays current with daily updates. Thank you!
Hey thank you for keep doing this flowise videos❤ . You and flowise AI together going to build so many dreams for us ❤ keep doing more flowise videos please . Btw bro a youtuber compared all tools and said flowise not good for production , the following was his take one flowise. do you agree with him ? Or whats your take on his comment plz do tell me His comment as follows: "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."
Hello @leonvanzyl, first of all, thank you for all videos about flowise AI :) you are the best ! Could you please explain how to scale it in production /live environnement ?
Hi Leon. Great video, many thanks! When you have time, can you please make a video on hooking up litellm with Flowise? The need for it comes from custom LLMs that you can not link through Ollama. (say you have a local cluster that provides LLM access via API). Thanks a lot!
Great video Leon. Quick question on latency. I see major degradation when it comes to agentic flow, from 2 sec to 20 secs. Is it normal? I used most of the steps shown by you, vector search is hardly 0.5-1 sec, so this 20 secs is contributed by other steps only.
Hey Leon, great video. You set the entry hurdle so low that you get good results straight away. Nevertheless, the trick is in the detail. I failed to build a chatbot that takes into account the code of a mermaid diagram to guide the user through a decision-making process. The prompt works fine when I run it with Llama 3.3 70B in OpenWebUI as a GUI, for example, but when I use it with a node in Flowise, the numerous brackets of the Mermaid code are misinterpreted as variables. Is there a way to avoid this? Can I mark the Mermaid code as a code area within the prompt?
Hi Leon, I've create an agent AI with the latest version of Flowise and the agent have OpenAI as a model and also have Supervisor and four Workers. When I run the agent, after the first worker, the Agent stop woking and send me the following message "Message Stopped". Please can you explain WHY and HOW can I solve the problem ? Thank you in advanve and thank you also for the great work !
If there is a requirement to work without internet connectivity, documents can be stored locally, and vectorization can be performed using locally running models like Ollama and Flowise.
Wow now thats super gold i just saw the title and duration of this video and super pumped , didnt watched it yet thats why i couldn't know that . Anyway thank you boss❤@@leonvanzyl
I am at 24:56 trying to run it but the chat window is only a blank page with nothing on it. digging into it for a cple of hours now but cant figure out whats wrong. guess some things never change strange thing also is whenever you click on the chat button there appear three different buttons. when I click that thing it automatically pulls me into the empty "chat". I can see the 3 buttons jump out of that chat button, but as soon as I can see them the page goes white :D
Thanks for the videos Leon are much appreciated, how about creating an assistant in Flowise that writes flows, an AI Coder to edit or generate Blueprint.json entirely? An AI model with knowledge of Github documentation and the “right instructions” could be a substantial step forward. Graziano
Hello Leon thanks a lot for your video, I have question because for my internship I have to create an agent workflow capable of evaluating the performance of an LLM, could you help me please ?
Can you build an appointment setting agent team which qualifies and books the call and if the lead is unqualified it starts lead nurturing by sending weekly emails
Wow good work 👏, keep doing more flowise master classes 👏. I got a question for you, is flowise good for building AI agents comparing to vectroshift another AI agent builders . I know flowise good for LLM orchestration but to build AI agents I wanna know is it as good and capable comapring to induatry standard no code tools like vectroshift etc or framework like Pydantic AI? And can we anyhow integrate framework like Pydantic AI in flowise AI agent building flow ? Plz answer this questions 🙏 and once again thank you❤
Thank you lean for your replay , Quick question about building a customer support chatbot 🤖 I'm planning to create a customer support AI agent, and I'm stuck on the pricing strategy. Could you help me out? 🙏 My Current Dilemma: What features should I include in a $200 base package (as you suggested keep 200 dollars as base variant price ) ? How many tokens/messages/characters should I limit? One-time fee or monthly subscription? 🤔 Is $200 enough for a solid customer support chatbot? I'm trying to create a competitive offering that provides real value to clients. Would love your expert insights on: Pricing structure Feature limits Knowledge base requirements Any advice would be super helpful! 🌟 Been following your channel and you're my go-to guru for tech insights. 💡 Thanks in advance!
Very good questions. I might need to create a dedicated video on monetizing these offerings. I don't charge the clients for usage. They need to bring their own API keys. These include LLM providers, SERP API, etc. The cost of document stores depends on the amount of files or time spent in connecting things like Confluence. Just decide on an hourly rate for yourself and charge time and material. The same goes for setting up tools and embedding the chat bot into their website.
This is great. About the Telegram convo, how can we ‘reset’ the chat ID where required? Eg say we don’t want the user to have weeks of convos with the same chat id where the memory includes multiple topics ? Eg maybe we want the chat to reset on each day, or via some sort of user request?
Thank you @leonvanzyl for sharing amazing flowise tutorial in such great detail for free. It means a lot to a learner like me. You are a AI expert with a kind heart;
I'm terrible at pricing for these projects. I tend to under charge, but you should keep in mind it's not about your time but about the value you bring to the client.
Hi Leon! It was an amazing video! Let me ask: now where n8n has advanced IA features, could we replace Flowise with n8n and build all stuff there? N8n is very nice, but I think it lacks a "Document Store" manager like in Flowise. What is your opinion? Thanks!
Yeah, the lack of a "document store" in n8n is painful. Something like Record Manager in FW has really been a game changer for me. Also, and more importantly, sequential agents in FW are awesome. I tried recreating some of the more advanced flows in n8n, but the lack of checkpoints and state management make it pretty much impossible.
hey mr leon, thanks to provide this again and again, can you help me ? is that possible for this idea use flowise to access my supbase DB and can automate query with only text i input like "can you find whos is login into this apps on monday" and the result will be show the data, thanks before mr 🙏
There is no "one size fits all" unfortunately, and I have videos on my channel for all of these platforms - including additional ones like VectorShift. However, I highly recommend Flowise and n8n.
Leon i heard that flowise is not recommend for production because it doesn't have the best logging/monitoring/ error handling, but good for fast prototyping only ? . Do you agree ? What u r take on this leon curios to know ? Can we integrate any 3rd party tool to solve this issue ? 🤔
PydanticAI is really cool. Let me know if you'd like to see a tutorial on it. Keep in mind that PydanticAI is a Python framework - not a no-code tool like FlowiseAI.
Leon i got a doubt , since flowise use langgraph for building agent can we use langgraph studio instead . What you think of langgraph studio vs flowise for agents.
This is random question but iam always curious tho , like you must have clients from all over the world right. How you receiving money from them ? What service you use to get paid from your international clients for your AI services ?
How much you charge from your client for a production ready flowise customer support chatbot . Just Give an aprox nunber so that we can know how the pricing goes for this product.
Depends on the complexity of things like the knowledge base and custom tools. I would also charge more if the bot needs to be integrated with Whatsapp - since the setup on Meta can be painful. Maybe set a base price of like $200 for the basic setup and increase the costs based on the tools and knowledge base. This can easily push the costs into the thousands.
Sir please Anyway to: chatbot reply when customer is not active for a certain time like: Sorry, the waiting time is over, if you still need help contact us again. Thank you for reaching out to us! ✨ and when the customer reply chatbot flow start from scratch. please help.
Your tutorial is very valuable. I learned a lot. Thank you. 🥰Can I ask you a question? When using n8n to connect to my Flowise instance, it seems that {{ $('Edit Fields').item.json.flowiseBaseURL }}/api/v1/document-store/loader/process is no longer functioning. It also appears to have been removed from the menu, and I couldn’t find any explanation in the documentation or updates.Has this feature been deprecated or replaced by a new API?
Hi Leon. Great video. I'm learning a lot. But I'm wondering if you know what I might be getting this error when I try to do the upsert process: Oh snap! The following error occured when loading this page. Status: 500 Error: documentStoreServices.insertIntoVectorStore - Error: documentStoreServices._insertIntoVectorStoreWorkerThread - Error: 403 Project proj_pAs8meLTsss5QIUyFaDNIBEg does not have access to model text-embedding-3-small . I have a paid OpenAI account. my API key works fine for the other nodes and tools. Not sure where to look for the root of the problem.
This video was a lot of effort to create. Subscribe to my channel to be notified of more FlowiseAI videos ⭐.
Let me know what other videos I should create - whether it's Flowise related or not.
I've been watching your videos,I am a big fan ,thank you and may Shiva bless you ❤
Yes! Flowise videos please. You are The best🎉
And as for Flowise, I would love to see a video on monetizing assistants and agents, e.g., using integration with Stripe and payment for tokens, flat fee or subscriptions.
Leon you are FREAKING AWESOME!! You are the goat 🙏🏼
Awesome work Leon, most appreciated! Champion!
I just came across your channel a few hours ago and I want to say a very big thank you to you. This is the first UA-cam channel in a while that is providing its users with practical knowledge and not this junk clickbait channels that are littered all over UA-cam adding zero value. That said Please make a detailed Flowise video showcasing how to create, store, fetch, and delete user data in Supabase based on each user's input on Telegram or WhatsApp. The video training that I am suggesting will get you a lot of viewership and subscribers since AI workers and agents will eventually render SaaS apps obsolete and when that time arrives users who don't want to communicate with apps via voice still need a UI by then to interact with the ai app and that UI will be either WhatsApp or Telegram.
Thank you for the amazing feedback
Leon hi! When you do not have a long video, I start to worry that you are tired of telling us amazing chips, but when a new video is released, and even more so a tutorial becomes clear what a colossal work has been done for us that we can create and be happy. I and my family are grateful for your efforts and if blogodorya them I can earn my first money, I will definitely be able to buy you a coffee machine)))) thank you!
Coffee machine sounds good! I turn caffeine into AI agents 😊.
Thank you for the amazing comment.
Leon, there are many videos on Flowise on UA-cam, but you are THE authority on it and your content is super valuable. Thank you for doing what you do.
Thank you!
What immense value! I have been following you since the first flowise video and you have helped me so much. I run my own AI Agency now creating AI solutions and I still come to your videos every now and then to learn something new. Keep up the great work!
This is incredible! I'd love to hear more about your agency. Please send me an email
Дякую! Дуже цікаві відео, хочу вивчити цей інструмент і стати єкспертом.
Thanks!
Thank you for the support ❤️
Wow Leon, just when you think your videos cannot get any better, Boom!! you shock the community again with amazingly easy to follow knowledge, great examples and real-life scenarios. I know we will meet one day because I want to shake the hand of someone who has inspired me to build great things. Thank you Leon! ⭐⭐⭐⭐⭐
Amazing comment! Thank you! Look forward to hearing about your successes when we meet 😊
Teşekkürler.
Thank you for the support!
Nobody breaks this stuff down and makes it accessible to almost anyone better than you, Leon. Excellent job.
Wow, thank you!
Thank you so much, Leon, for your outstanding effort. Your wisdom has been an incredible benefit to me, and I feel like the greatest recipient of it. Thank You again.
Great work! I love how your channel is my number 1 place to visit for anything Flowise related, thank you!
Leon, your videos are an absolute game-changer! Your way of breaking down complex AI concepts makes it so much easier to understand and keeps me motivated on my learning journey. While I’m not quite there yet, your content makes me believe it’s possible to start building some amazing agents and web apps soon. Thank you for the consistent value you bring!
Thank you for the amazing feedback.
Have confidence in yourself. You're able to do this.
Welkom terug Leon, en beste wense vir 2025! Ek het baie uitgesien na nog ’n FW-video - uitstekend, baie dankie
Voorspoed vir die nuwejaar!
Any chance to expand your videos for n8n?
Absolutely!
@@leonvanzylPlease. Will love to see more n8n too if you will
Spend 15 mins yet and looks amazing , thanks
Amazingly well-structured, packed with practical info, and concise video, thank you!
You're welcome
Lovely well structured tutorial
Thank you
Leon, I told that person who previously said Flowise isn't production-ready due to logging, monitoring, and error handling issues. Now, he's changed his stance, but still mentions some concerns that I don't fully understand. What should I tell him, Leon? You're my expert in this, and I'm fully supporting Flowise!
He said:
'I didn't know Flowise integrates with LangSmith, which addresses a major limitation. LangSmith helps with logging and monitoring, but not error handling. It's still a drawback to rely on a specific tool (LangSmith) and not have easy customization options.
However, these downsides don't mean Flowise isn't worth using. I think it's a phenomenal platform!'
LOL, don't you just love debating people in the comment section 😂.
I'm not sure what he means by error logging then.
1) Flowise logs any technical errors in the console which you can view in the logs if you're using something like Render. Flowise cloud has a dedicated menu options for these logs as well - making it super simple to view any errors.
2) Langsmith (and the other platforms that FW can integrate with) do offer error logging as well if the Langchain Runnables failed for any reason (including issues with tools or LLM providers).
Great job, Leon! It's a pleasure to watch your work. I can see that you've put a lot of time into it. Best regards!
Thank you! Cheers!
You are doing a grate job !!! Thank you!!! I hope to see flows with API loaders and agents integrations) that would be very usefull) for example integrations between different platforms through their REST APIs) Thank you ones again!
Happy new year brother.. Great work again 👏🏾👏🏾
Happy new year!
Leon, thank you for your prompt response! I'm eager to learn from your tutorial on Pydantic AI. Could you also provide insights on why Pydantic AI stands out and how it compares to no-code tools like Flowise in terms of production deployment? Your expertise in showcasing top-tier AI automation and agent products is truly impressive, and I'm excited to gain knowledge about production-ready AI applications and AI Agents. Even if transitioning from pure no-code to minimum code, I'm interested in learning how to achieve what no-code tools can't achieve with minimal coding from you!
Hi Leon, thanks for the videos you are the, it would be very helpful to create a video with a guide on how to connect Flowise to Make!
I actually have several videos on integrating FW with Make 😊.
Since they've been dodging paying creators affiliate commission, I moved all videos over to n8n.
항상 감사드립니다. 새해 복 많이 받으세요 :)
Hi Leon. A very happy new year. No one makes FlowiseAI contents better than you. I think we will need your help with a cheat sheet of nodes based on specific use cases. There are so many nodes available. Sometimes, it becomes overwhelming. I think we will start to see community contribution towards node documentation pages as well.
This video covers all the nodes you need to know 👍. Seriously, I know there are many nodes, but with the nodes in this video, you can achieve anything.
Excellent as always Leon. A very good idea for an update video and as detailed as this one. We appreciate your dedication
Awesome work! Already subscribed, but shared it among friends who are also busy with their Ai projects. Hopefully more ppl will share this. Amazing work!
Much appreciated!
Flowise is definitely a great tool for building complex scenarios. Unfortunately, it's often hard to understand everything you can do and all the capabilities offered by every component, simply because of the lack of documentation. For example, I was not able to use or even understand how to set up the additional parameters for the chat prompt template, and I have many other examples like this. I hope the documentation will improve to demonstrate the quality of the product and allow everyone to build advanced flows.
love it. subscribed. looking forward to more flowise tutorials and courses. great course
Thank you! More Flowise is coming!
Thank you! It's a great way to kick off the new year. I usually code simple Python apps on my own, but I use Flowise for fast prototyping. It's a very useful tool! :)
Most anticipated video of the year! 😊
Hope you enjoyed it!
INSANE Video Leon. You are Mr. Flowise 💪 What a great job 👍 Thank you very much. Yes please MORE from Flowise!
You got it!
Amazing start for 2025🎉 Thanks Leon
Thank you, I hope you enjoy the course!
Happy new year Leon, will dig in your tuto this 2025, thanks for your sharing knowledge
Happy new year!
Huge thanks! I learned a lot thanks to your videos. JSON downloads are great to fix mistakes in my flows. I highly appreciate you are sharing them!
You're welcome 😊
This is excellent!
Can you please create one on adding Figma data to the Document Store & using it?
thanks for this. perfect timing for my needs. appreciate all your work and efforts!
You're welcome!
Awesome video Leon, Thank you, I am sure it took a fair amount of effort
Yeah, it was 😅
I legitimately felt like I should have paid for this. Thank you!
I hinestly planned to make this a paid course on Udemy, but I decided to upload it for free for some reason 😅
@leonvanzyl it builds your credibility with me.
Leon! noce to see your tutorials, you helped a lot to my students!!! keep going. happy new year 🎉
That's awesome! Glad I could help
Thank U So much Leon 🦁You are my favorite AI UA-camr! ...It would be great how we can interact with databases, a Classic CRUD for example. Greetings from México...
Greetings from South Africa 🇿🇦.
Thank you for the feedback and suggestions
Thanks for this great work! I’d like to ask if you could create a video tutorial on using Firecrawl’s API to scrape a website with more than 200 tabs and subpages. To update a chatbot knowledge base every 24 hours. If you know of any other method that might be better suited for this scale and frequency, we’d love to hear about that as well. The key for me is ensuring the knowledge base stays current with daily updates. Thank you!
Thank you for dropping resources
You're welcome 😊
Great video.
Thanks!
Hey thank you for keep doing this flowise videos❤ . You and flowise AI together going to build so many dreams for us ❤ keep doing more flowise videos please . Btw bro a youtuber compared all tools and said flowise not good for production , the following was his take one flowise. do you agree with him ? Or whats your take on his comment plz do tell me
His comment as follows:
"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."
Sounds like a very outdated take on FW. I think that I addressed most of those concerns in my intro and throughout the video.
Hello @leonvanzyl, first of all, thank you for all videos about flowise AI :) you are the best ! Could you please explain how to scale it in production /live environnement ?
Hi Leon. Great video, many thanks!
When you have time, can you please make a video on hooking up litellm with Flowise? The need for it comes from custom LLMs that you can not link through Ollama. (say you have a local cluster that provides LLM access via API). Thanks a lot!
Thank you. Will look into Lite LLM.
Great video Leon. Quick question on latency. I see major degradation when it comes to agentic flow, from 2 sec to 20 secs. Is it normal? I used most of the steps shown by you, vector search is hardly 0.5-1 sec, so this 20 secs is contributed by other steps only.
Hey Leon, great video. You set the entry hurdle so low that you get good results straight away. Nevertheless, the trick is in the detail. I failed to build a chatbot that takes into account the code of a mermaid diagram to guide the user through a decision-making process. The prompt works fine when I run it with Llama 3.3 70B in OpenWebUI as a GUI, for example, but when I use it with a node in Flowise, the numerous brackets of the Mermaid code are misinterpreted as variables. Is there a way to avoid this? Can I mark the Mermaid code as a code area within the prompt?
I can't thank you enough. 👏👏👏
You're welcome 😊
Hi Leon,
I've create an agent AI with the latest version of Flowise and the agent have OpenAI as a model and also have Supervisor and four Workers. When I run the agent, after the first worker, the Agent stop woking and send me the following message "Message Stopped". Please can you explain WHY and HOW can I solve the problem ?
Thank you in advanve and thank you also for the great work !
Great training on an excellent platform! Would you consider creating a training session focused on graph flow with Neo4j?
Working on it
If there is a requirement to work without internet connectivity, documents can be stored locally, and vectorization can be performed using locally running models like Ollama and Flowise.
Absolutely!
I have videos on that as well.
@@leonvanzyl please share the link of that video
@@leonvanzyl link of that video?
@BilalKundi-lm5zz ua-cam.com/video/lJOZiRoZNJw/v-deo.html
Can you create some kind of membership to share advanced flows? like real projects and stuff. We will sign up super quick. Thank you for all you do
Now this is gold ♥️😍😍😍... do same thing for N8N and at the end do for N8N + Flowise Master class boss ❤😮
Thank you.
You should know that the last project in this video does include n8n + Flowise! Enjoy
Wow now thats super gold i just saw the title and duration of this video and super pumped , didnt watched it yet thats why i couldn't know that . Anyway thank you boss❤@@leonvanzyl
The best 🎉
Thanks
brilliant
I am at 24:56 trying to run it but the chat window is only a blank page with nothing on it. digging into it for a cple of hours now but cant figure out whats wrong. guess some things never change
strange thing also is whenever you click on the chat button there appear three different buttons. when I click that thing it automatically pulls me into the empty "chat". I can see the 3 buttons jump out of that chat button, but as soon as I can see them the page goes white :D
running it in docker works fine though
Thanks for the videos Leon are much appreciated, how about creating an assistant in Flowise that writes flows, an AI Coder to edit or generate Blueprint.json entirely? An AI model with knowledge of Github documentation and the “right instructions” could be a substantial step forward. Graziano
Ooh, that sounds awesome...
Hello Leon thanks a lot for your video, I have question because for my internship I have to create an agent workflow capable of evaluating the performance of an LLM, could you help me please ?
Can you build an appointment setting agent team which qualifies and books the call and if the lead is unqualified it starts lead nurturing by sending weekly emails
Leon i saw ina comment u said langgraph studio and flowise solve different prombkems in building agents ? What are they plz do explain . Thanks
Wow good work 👏, keep doing more flowise master classes 👏. I got a question for you, is flowise good for building AI agents comparing to vectroshift another AI agent builders . I know flowise good for LLM orchestration but to build AI agents I wanna know is it as good and capable comapring to induatry standard no code tools like vectroshift etc or framework like Pydantic AI? And can we anyhow integrate framework like Pydantic AI in flowise AI agent building flow ? Plz answer this questions 🙏 and once again thank you❤
Flowise is extremely good at builder agent flows. I'd recommend watching the sections on Agentflows
Thank you lean for your replay ,
Quick question about building a customer support chatbot 🤖
I'm planning to create a customer support AI agent, and I'm stuck on the pricing strategy. Could you help me out? 🙏
My Current Dilemma:
What features should I include in a $200 base package (as you suggested keep 200 dollars as base variant price ) ?
How many tokens/messages/characters should I limit?
One-time fee or monthly subscription? 🤔
Is $200 enough for a solid customer support chatbot?
I'm trying to create a competitive offering that provides real value to clients. Would love your expert insights on:
Pricing structure
Feature limits
Knowledge base requirements
Any advice would be super helpful! 🌟 Been following your channel and you're my go-to guru for tech insights. 💡
Thanks in advance!
Very good questions. I might need to create a dedicated video on monetizing these offerings.
I don't charge the clients for usage. They need to bring their own API keys. These include LLM providers, SERP API, etc.
The cost of document stores depends on the amount of files or time spent in connecting things like Confluence. Just decide on an hourly rate for yourself and charge time and material.
The same goes for setting up tools and embedding the chat bot into their website.
@@leonvanzylthsnks boss..plz do a dedicated video on this ❤
Thank you my Friend
You're welcome
This is great.
About the Telegram convo, how can we ‘reset’ the chat ID where required?
Eg say we don’t want the user to have weeks of convos with the same chat id where the memory includes multiple topics ? Eg maybe we want the chat to reset on each day, or via some sort of user request?
In Telegram, I think you need to clear / delete the conversation
Thank you @leonvanzyl for sharing amazing flowise tutorial in such great detail for free.
It means a lot to a learner like me. You are a AI expert with a kind heart;
Thank you 😊
Hi leon
what should be an ideal cost/charges for deploying ai chatbot through flowiseAI for a client (excluding everything else)
I'm terrible at pricing for these projects. I tend to under charge, but you should keep in mind it's not about your time but about the value you bring to the client.
나중에는 Flowise와 N8N이 결합된 모델도 기대합니다. 감사합니다 :)
Hi Leon! It was an amazing video! Let me ask: now where n8n has advanced IA features, could we replace Flowise with n8n and build all stuff there? N8n is very nice, but I think it lacks a "Document Store" manager like in Flowise. What is your opinion? Thanks!
Yeah, the lack of a "document store" in n8n is painful. Something like Record Manager in FW has really been a game changer for me.
Also, and more importantly, sequential agents in FW are awesome. I tried recreating some of the more advanced flows in n8n, but the lack of checkpoints and state management make it pretty much impossible.
Thanks Leon!
hey mr leon, thanks to provide this again and again, can you help me ? is that possible for this idea use flowise to access my supbase DB and can automate query with only text i input like "can you find whos is login into this apps on monday" and the result will be show the data, thanks before mr 🙏
I'll see if I can create a video on this. I've been meaning to create a SQL agent for a while now
Make a video about how to monetize your own created Agents
thanks ,for ai apps ,agents,flows,products what is better n8n,flowise,langflow?
There is no "one size fits all" unfortunately, and I have videos on my channel for all of these platforms - including additional ones like VectorShift.
However, I highly recommend Flowise and n8n.
Leon i heard that flowise is not recommend for production
because it doesn't have the best logging/monitoring/
error handling, but good for fast prototyping only ? . Do you agree ? What u r take on this leon curios to know ? Can we integrate any 3rd party tool to solve this issue ? 🤔
I don't agree at all - as shown in this video FW offers integration with several analytics tools - like LangSmith and others.
Leon, is it possible to do the last step, i.e. connecting Flowise to Telegram if I have Flowise self hosted free version on an external server?
Absolutely!
Hey leon what you think of Pydantic Ai framework and can we achive same thing in flowise .
PydanticAI is really cool. Let me know if you'd like to see a tutorial on it.
Keep in mind that PydanticAI is a Python framework - not a no-code tool like FlowiseAI.
Leon i got a doubt , since flowise use langgraph for building agent can we use langgraph studio instead . What you think of langgraph studio vs flowise for agents.
They're very different.
I use Langgraph studio a lot and feel these two platforms solve different problems.
@@leonvanzyl which one is better and what problems each solves ?
I have a small problem, i cant see anything in the marketplace, on my machine.
Hit the refresh button, its a glitch
This is random question but iam always curious tho , like you must have clients from all over the world right. How you receiving money from them ? What service you use to get paid from your international clients for your AI services ?
Paypal. Would have preferred Stripe, but it's not available in South Africa 😢
Can you activate automated dubbing ?
You mean on the video itself? It is activated. I think it takes YT a while to process the dubbing
@@leonvanzyl Thx !
Leon, How can I improve my bot response time it takes approximately 10-15 seconds to respond.
Which model are you using?
Try using a smaller model like GPT-4o
@leonvanzyl yes I am using the same, Bot is made of Open ai assistant and custom tools.
i have trying phidata for agent, always comeback to flowise 😅
Flowise is not working with nodes 23, pls fix t
I'm not the developer 😄.
It's working for me though.
It gives me an error: Cannot find module 'langchainhub'
Haven't seen that error myself. What are the steps to recreate the issue?
@@leonvanzyl It appears after px flowise start command is run. I've reinstalled Node.js but nothing changes. I have win11.
C:\Users\Lucru>npx flowise start
» ModuleLoadError: [MODULE_NOT_FOUND] require failed to load
» C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise\dist\commands\start.js: Cannot find module 'langchainhub'
» Require stack:
» - C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise\dist\index.js
» - C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise\dist\commands\start.js
» - C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise
ode_modules\@oclif\core\lib\module-loader.js
» - C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise
ode_modules\@oclif\core\lib\config\plugin.js
» - C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise
ode_modules\@oclif\core\lib\config\config.js
» - C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise
ode_modules\@oclif\core\lib\config\index.js
» - C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise
ode_modules\@oclif\core\lib\command.js
» - C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise
ode_modules\@oclif\core\lib\index.js
» - C:\Users\Lucru\AppData\Roaming
pm
ode_modules\flowise\bin
un
» Code: MODULE_NOT_FOUND
How much you charge from your client for a production ready flowise customer support chatbot . Just Give an aprox nunber so that we can know how the pricing goes for this product.
Depends on the complexity of things like the knowledge base and custom tools.
I would also charge more if the bot needs to be integrated with Whatsapp - since the setup on Meta can be painful.
Maybe set a base price of like $200 for the basic setup and increase the costs based on the tools and knowledge base. This can easily push the costs into the thousands.
This playlist seems to be inverted, the first video is last and the last one is first.
Thanks for letting me know.
Instead of telegram can i do the same for whatsapp , if yes how though and will i get banned from meta for doing it ? Since meta is too restrictive.
I already have a Whatsapp tutorial on my channel
Btw Leon your website chatbot is not working 😢
That's the rate limiting kicking in. People spamming my poor bot since the release of this video 😅
Sir please Anyway to: chatbot reply when customer is not active for a certain time like: Sorry, the waiting time is over, if you still need help contact us again.
Thank you for reaching out to us! ✨ and when the customer reply chatbot flow start from scratch. please help.
What's the use case? Why would you want to do that?
@@leonvanzyl In WhatsApp cloud api chatbot - I want to terminate the session if user not active with the chatbot for a while.
@@leonvanzyl I want to close the chatbot session after some time when user is unactive. and restart the Chabot when user reply again.
Your tutorial is very valuable. I learned a lot. Thank you. 🥰Can I ask you a question? When using n8n to connect to my Flowise instance, it seems that {{ $('Edit Fields').item.json.flowiseBaseURL }}/api/v1/document-store/loader/process is no longer functioning. It also appears to have been removed from the menu, and I couldn’t find any explanation in the documentation or updates.Has this feature been deprecated or replaced by a new API?
Follow the steps in this video rather
Hi Leon. Great video. I'm learning a lot. But I'm wondering if you know what I might be getting this error when I try to do the upsert process:
Oh snap! The following error occured when loading this page. Status: 500 Error: documentStoreServices.insertIntoVectorStore - Error: documentStoreServices._insertIntoVectorStoreWorkerThread - Error: 403 Project proj_pAs8meLTsss5QIUyFaDNIBEg does not have access to model text-embedding-3-small .
I have a paid OpenAI account. my API key works fine for the other nodes and tools. Not sure where to look for the root of the problem.
@leonvanzyl i have asked if you can create please an chatbot in flowise that supports Arabic with Hugging Face instead of Ollama please ....
Already have a video on Hughingface and FW 😎
ua-cam.com/video/rGehI5PNP2o/v-deo.html