Creating Chatflows & LLM Chains - FlowiseAI Tutorial #2

#flowiseai #flowise #openai #langchain
LLM Chains might be the simplest chain, but they are extremely powerful.
In this video we will have a look at creating our first chatflow using an LLM Chain and OpenAI model.
☕ Buy me a coffee:
www.buymeacoffee.com/leonvanzyl
📑 Useful Links:
OpenAI API Keys: platform.openai.com/api-keys
💬 Chat with Like-Minded Individuals on Discord:
/ discord
🧠 I can build your chatbots for you!
www.cognaitiv.ai
🕒 TIMESTAMPS:
00:00 - Introduction to LLM Chains
02:16 - Adding an LLM Chain
03:52 - Adding a Model
04:42 - OpenAI API Key
05:07 - Viewing Credentials
06:39 - Adding Prompt Template

[@leonvanzyl]

Thank you all very much for the love and support. Let's see if we can hit 1000 likes on this video! Hope you enjoy 😎.

[@fernandogg152]

Hi bro, thanks for the videos. Luv u so much

[@olivierhagoort8083]

Awesome video's! look forward to more Flows content!

[@Organizational_Psychology_PhD]

Love your Flowise tutorials! You helped me get up and running with my own AI bots back in October and now I’m far ahead of all of my peers. You rock!

[@leonvanzyl]

That is just amazing to hear!

[@wgaitanartunduaga]

Great tutorial! Your step-by-step explanation makes setting up Flow-wise a breeze. Looking forward to diving deeper into the integration options and exploring the advanced features in the upcoming videos. Thanks for the clear guidance!

[@leonvanzyl]

🙏

[@ChatbotWhizAmir]

Awesome Leo! Thanks and Please keep up on the great work!

[@leonvanzyl]

Thank you

[@PepettosClub]

the best tutorial series I've ever seen

[@leonvanzyl]

Wow, thank you!

[@altonjones5271]

Thanks Leon - took your advice on Pinecone and finally have a working RAG app with Flowise! Thank you once again!

[@leonvanzyl]

Whoohoo!

[@ward_jl]

Love your videos, Leon. Looking forward to learning how to create more 'intelligent' agents by integrating functions!

[@leonvanzyl]

Thank you! We'll definitely cover agents and tools in this series. Stick around 😁

[@anilrajshinde7062]

Your all tutorials are really useful.

[@anilrajshinde7062]

I want to create RAG application using Flowise. I want to upload pdf without canvas or chatflow place. How can I do that? give your valuable inputs

[@leonvanzyl]

Thank you 🙏

[@user-nd1gb1pi3f]

Really Like your this efforts

[@leonvanzyl]

You're welcome 🤗

[@cmca1607]

Hola Leon, quiero agradecerte por tu tutorial es muy bueno, espero que puedas entender este comentario o más bien pregunta, aunque esté en español. Cuando te refieres a "connect credencial" del nodo OpenAI, te aparece una venta con la cual seleccionas una altenativa, sin embargo en mi caso no me aparece ninguna ventana. Ni en el ejemplo, ni en el menú principal. Tienes alguna idea de por qué se puede producir esto? Gracias

[@digitald74]

thanks for your tutorial.. watching one by one. do you know if the chat client supports buttons? i know that you can have starter prompts but i wonder if we can manipulate the output from the llm to show buttons for different use cases. like talk to sales / talk to support.

[@leonvanzyl]

Flowise doesn't support buttons.. it kind of defeats the purpose of using NLU, and is more of a "traditional" Chatbot approach. You could always use something like Botpress or Voiceflow instead 👍

[@RenatoCadecaro]

🎯 Key Takeaways for quick navigation: 00:00 *🤖 Flowwise AI é uma ferramenta poderosa para criar aplicações de IA sem codificação.* 00:28 *🛠️ Flowwise permite criar chatbots e agentes de IA de forma fácil com sua interface de arrastar e soltar.* 01:00 *🌐 Configurar o Flowwise localmente é simples, e ele pode ser integrado a sites ou aplicativos usando pontos de inserção ou API.* 01:29 *📦 Instalar o Flowwise globalmente ou em uma pasta específica é fácil com comandos simples.* 02:27 *💻 Atualizar e desinstalar o Flowwise também são processos diretos.* 03:27 *📂 Instalar o Flowwise em uma pasta própria abre recursos avançados.* 04:25 *🔄 Automatize a instalação do Flowwise com um script fornecido, simplificando o processo.* 05:55 *🚀 Após a instalação, explore a interface do Flowwise para criar e gerenciar fluxos de conversa e utilizar modelos e ferramentas pré-criados.* 07:52 *💬 Os modelos no Flowwise podem ser salvos e gerenciados facilmente, permitindo a criação rápida de novos fluxos de conversa.* 08:21 *🛠️ Ferramentas personalizadas podem ser atribuídas às aplicações de IA no Flowwise, oferecendo uma ampla gama de funcionalidades adicionais.* 08:51 *🔑 As credenciais, incluindo chaves de API, podem ser gerenciadas centralmente no Flowwise, facilitando a integração com serviços externos e a segurança dos dados.* Made with HARPA AI

[@lauraramon8870]

Thanks for your videos, you explain very well! Can you please make a tutorial to create a AI chatbot to chat with my own pdf documents with open source models, like Llama2 hosted in HuggingFace, for example? I mean, no Open AI API keys. I think that is very interesting theme. Thanks!

[@leonvanzyl]

Great idea. Will be sure to include that once we get to the open source models section.

[@cmca1607]

Hello Leon, I want to thank you for your tutorial is very good, I hope you can understand this comment or rather question, although it is in Spanish. When you refer to "connect credential" of the OpenAI node, you see a window with which you select an alternative, however in my case I do not see any window. Neither in the example, nor in the main menu. Do you have any idea why this could happen? Thanks

[@leonvanzyl]

Perhaps you have an old instance of Flowise.

[@ward_jl]

@leonvanzyl, do you have any advice on how to update Flowise on Render without losing my current projects? I am running it there as demonstrated in one of your older videos. Would appreciate your feedback, and follow along with your videos using an updated version of Flowise!

[@leonvanzyl]

I will cover this in the series, but all you have to do is to go to YOUR Flowise repo and click "sync changes". This will automatically update Flowise on Render and you will not lose your Flows. Granted you're not on the free plan of course.

[@olivierhagoort8083]

Hi Leon, I'm trying to build a website with a Chatgpt like interface. So far I have managed to create a chatflow based on OpenAI Assistant with tools like web browser etc. I have also seen your video on creating long-term memory but I have read that it is not necessary to do so for OpenAI Assistants. My question to you is the following: I want my chatbot interface to showcase a list of chat history (userID specific) which they can visit and go back to earlier conversations, just like in chatgpt. I know I can see the message history but I want the user to acces his/hers as well. Is this possible through Flowise? If so, how do I do it?

[@leonvanzyl]

Hey there. This goes beyond the scope of Flowise, but is definitely possible. You can use Flowise to interact with the assistant using the Flowise API endpoint, but the rest of the points you made will have to be developed outside of Flowise. I assume your app involves user authentication. You could then store the threads / conversations per user, and then use the Assistants API to fetch the past messages, etc. My agency can assist you will this.

[@Aidev7876]

I know you kind of answered that but I'm debating with myself whether deploying flowise on aws and using the low code, or just langchain plus something like fastAPI to expose it. I'm a developer, yes. Can you throw another bone here?

[@leonvanzyl]

I feel your pain. I'm in the same boat in my profession 😁. Flowise is really well developed.. because it's Open Source, the contributors have done an incredible job at following best practices and building a performant tool. Coding everything from scratch and applying good principles, exception handling, etc. can be time consuming, and in a sense, reinventing the wheel. I've used Flowise to develop client solutions, and I've also used it to prototype Langchain applications which I eventually turned into a coded solution.

[@Aidev7876]

@leonvanzyl can flowise be exposed to a rest API? I've a customer who will consume my API, passing message and user ID for example. My API will do the AI stuff and respond. Obviously it needs to support memory for follow up questions. Can flowise do all that?

[@fernandogg152]

Leon, how is going? I have a question When u gonna upload the next episodes of the serie? Thank u

[@leonvanzyl]

Hahaha, working on them 😂

[@dr.robsonchamon]

Hey Leon! First, thnks for the videos!! I’m having trouble maintaining my flowise station running on render, it constantly keeps running down.. its a paid instance. Can you refer me to a tutorial of yourself that might help me set up correctly só i van have a reliable chatbot? Thank you

[@leonvanzyl]

Check out the deployment video on my channel. I will also be creating a refresh of that video in this series. The older video will still work though.

[@user-dc8hw3hz5e]

In the latest 1.8.0, I could not locate the Analyse Chatflow under the setting tab, Even after creating and running the chat bot. Anything, I am missing out ?

[@leonvanzyl]

It's moved to settings / config / analyze

[@florentflote]

[@leonvanzyl]

❤️

[@b0stj4n]

Hello, I get the "429 You exceeded your current quota error". Do you know a solution to that?

[@leonvanzyl]

Hey there. That means you've exceeded your OpenAI credit or your free credits have expired.

[@b0stj4n]

@@leonvanzyl Yes, you are right. I checked the Usage and my balance expired last year in April. Guess I need to pay for it or use a new account :D

[@hassanahmad1483]

My flowise is keep on downloading...not completing

[@leonvanzyl]

First time hearing about that. Could be your network or PC settings. Also, installation can easily take several minutes to complete. Or try one of the other methods. You could also just deploy it to Render (there's a video on my channel for Render deployment).

[@hassanahmad1483]

@@leonvanzyl Thank You...I will check the other method as you mentioned...

[@hassanahmad1483]

@@leonvanzyl installation completed but when i try to run flowise...getting error : Error: Cannot find module 'C:\Users\AppData\Roaming pm ode_modules\flowise\ bin un' ←[90m at Function.Module._resolveFilename (internal/modules/cjs/loader.js:962 :15)←[39m ←[90m at Function.Module._load (internal/modules/cjs/loader.js:838:27)←[39m ←[90m at Function.executeUserEntryPoint [as runMain] (internal/modules/run_ma in.js:71:12)←[39m ←[90m at internal/main/run_main_module.js:17:47←[39m { code: ←[32m'MODULE_NOT_FOUND'←[39m, requireStack: []