IMPORTANT: Forgot to mention this, but you need to add disk persistence in Render otherwise projects disappear if it crashes or reboots. Watch this video at 8:25 - ruclips.net/video/Fxyc6-frgrI/видео.htmlsi=GP02-S-PtFlRNg7z&t=505 👉 Want an AI chatbot built for your business? Visit omnifusion.ai
Thanks! Finally, a practical RAG example for Flowise has emerged. Surprisingly, Flowise hasn't provided a serious typical RAG example before. The official examples using an in-memory vector store are entirely unhelpful.
No, you just need to make sure your hosting (in this case Render) supports the volume you are trying to do, and if necessary scale up hosting resources. Also OpenAI needs to be funded so that you don't hit your usage limits. But you need a lot of conversations to hit that.
With Flowise, anybody can create and implement AI chatbots. For those wishing to incorporate AI into their company, it's a vital tool. #Flowise #AIChatbots #TechSimplified
Hi, the tutorial is awesome but can you tell me, how I can customize the colors of the links in the bot chatbubble? I have a dark chat bubble and the links are kinda black/dark blue. Any idea? Greetings from Austria Alex
You can change the CSS in the code of the Flowise chatbot or you can integrate it into Voiceflow for example if you want to use a different frontend to customize with no code.
If I deploy using the starter pack (512 mb ram and 0.5 cpu), how many chats will that set up be able to handle? Can you please give me ballpark estimations on: 1 - The max number of concurrent chats that it can handle and 2 - The max number of chats it can handle in a month assuming none of the chats are concurrent?
I'd say the starter pack should not be used in production. 512mb RAM is very low. We personally use 2GB RAM and it can handle multiple client projects with hundreds of msgs a day. Might also be worth looking into Digital Ocean as an alternative to render. 2. If the chats are not concurrent it really doesnt matter bc the RAM is only affected by whatever is happening at a given time.
too much info packed in a single video. Works beutifully,. But the chatbot hallucinates some of the answers even after instructions not to hallucinate.
@@OmniFusionAI thanks for the response. Did that, temp is 0.3 but still bring up unwanted stuff. Also replaced buffer window with a upstash redis memory. This helped in increasing the context window . But hallucination has not stopped :(
IMPORTANT: Forgot to mention this, but you need to add disk persistence in Render otherwise projects disappear if it crashes or reboots. Watch this video at 8:25 - ruclips.net/video/Fxyc6-frgrI/видео.htmlsi=GP02-S-PtFlRNg7z&t=505
👉 Want an AI chatbot built for your business? Visit omnifusion.ai
This literally just happened to me... Fk my life
This video is heavily underrated, I literally used it to crash course Flowise for a project.
Awesome job guys 🙏🏻
Glad you enjoyed it Mark, appreciate your feedback G!
Thanks! Finally, a practical RAG example for Flowise has emerged. Surprisingly, Flowise hasn't provided a serious typical RAG example before. The official examples using an in-memory vector store are entirely unhelpful.
Glad it helped!
Fantastic video! What do you think about Langflow builder? Is it better than Flowise or not?
Limited experience with it but I think flowise is top of the line.
🐐🔥 AI Goats
How long does it take to get that invitation from flowise?
If you watch the video you will see that you can self host - which has no wait list
How long does Flowise take to grant access since they have put me on a waitlist
The waitlist is only for their hosted version. Watch this video, I show how to self host from github.
Hey just a question, im new to these stuff. Is there a limit for how many messages or conversations the bot can have?
No, you just need to make sure your hosting (in this case Render) supports the volume you are trying to do, and if necessary scale up hosting resources. Also OpenAI needs to be funded so that you don't hit your usage limits. But you need a lot of conversations to hit that.
With Flowise, anybody can create and implement AI chatbots. For those wishing to incorporate AI into their company, it's a vital tool. #Flowise #AIChatbots #TechSimplified
True
How about a multimodal text vision text chat bot
Currently don’t have a tutorial for that but definitely possible
Hi, the tutorial is awesome but can you tell me, how I can customize the colors of the links in the bot chatbubble? I have a dark chat bubble and the links are kinda black/dark blue. Any idea?
Greetings from Austria
Alex
You can change the CSS in the code of the Flowise chatbot or you can integrate it into Voiceflow for example if you want to use a different frontend to customize with no code.
How do you get flowise to run off of your IP address rather than localhost?
I would deploy it on Render as shown in the video to make things easy
If I deploy using the starter pack (512 mb ram and 0.5 cpu), how many chats will that set up be able to handle?
Can you please give me ballpark estimations on:
1 - The max number of concurrent chats that it can handle and
2 - The max number of chats it can handle in a month assuming none of the chats are concurrent?
I'd say the starter pack should not be used in production. 512mb RAM is very low. We personally use 2GB RAM and it can handle multiple client projects with hundreds of msgs a day. Might also be worth looking into Digital Ocean as an alternative to render.
2. If the chats are not concurrent it really doesnt matter bc the RAM is only affected by whatever is happening at a given time.
@@OmniFusionAI That makes sense, thank you!
Hi I copy everything step by step but when I try to chat with agent I get "Error: Forbidden"?
Some api key may not be setup correctly. Either OpenAI or qdrant
@@OmniFusionAI Yes. all working now thanks for the fast response 👍
I see a "Message Stopped" error - any idea how to solve?
Double check qdrant and openai api keys. Personally haven’t experienced that error
too much info packed in a single video. Works beutifully,. But the chatbot hallucinates some of the answers even after instructions not to hallucinate.
I find a lot of people prefer a single video to learn compared to finding lots of small ones. If it hallucinates, turn down the temperature 🤝
@@OmniFusionAI thanks for the response. Did that, temp is 0.3 but still bring up unwanted stuff. Also replaced buffer window with a upstash redis memory. This helped in increasing the context window . But hallucination has not stopped :(