No Code Tools to use Gen AI

Gen AI, Rags, containers…

But what about the use cases?

RAG 101

My Fav RAGs Post

Docker Config Files Repository

Tinkering with Docker Services

The AI tech side might be perceived as chaotic…but you can keep it close to apps you are familiar, like telegram bots

Flowise AI

Flowise provides modular building blocks for you to build any agentic systems, from simple compositional workflows to autonomous agents

Build AI Agents, Visually

• Flowise AI, a groundbreaking Free platform that is making AI more accessible than ever before.
  • The Flowise AI Official Page
  • The Flowise AI Source Code at Github
    • License: Apache v2 ✅

Flowise x Streamlit

• Developer Friendly

  • API, SDK, Embed
  • Extend and integrate to your applications using APIs, SDK and Embedded Chat

• APIs

  • Embedded Widget
  • Typescript & Python SDK

• https://docs.flowiseai.com/integrations/3rd-party-platform-integration/streamlit

Embedded Flowise Chat

• https://docs.flowiseai.com/using-flowise/embed

• https://github.com/FlowiseAI/FlowiseChatEmbed

• https://github.com/ZynthCode/flowise-ai-with-python

LangFlow

Langflow is designed to redefine the boundaries of AI development in experimentation and real-world scenarios.

Its a lowcode visual framework for building multi-agent and RAG applications.

It’s open-source, Python-powered, fully customizable, model and vector store agnostic.

It is based on Python, which makes it ready to have many integrations:

• LangFlow:

  • The Flowise AI Official Page
  • The Flowise AI Source Code at Github
    • License: MIT

• https://hub.docker.com/r/langflowai/langflow

⛓️ Langflow is a visual framework for building multi-agent and RAG applications. It’s open-source, Python-powered, fully customizable, model and vector store agnostic.

python -m venv langflow #create the venv (python3 if you are in mac/linux)

langflow\Scripts\activate #activate venv (windows)
source langflow/bin/activate #(linux)

pip install langflow #https://pypi.org/project/langflow/  #--pre --force-reinstall 

#langflow run

#deactivate #when you are done

Acces LangFlow at localhost:7860

• https://github.com/leonvanzyl/langflow-tutorial-2024/blob/master/run.sh

Thanks again Leon Van Zyl

#version: '3.8'

services:
  langflow:
    image: langflowai/langflow:latest
    ports:
      - "7860:7860"
    stdin_open: true # equivalent to -i
    tty: true        # equivalent to -t
    restart: unless-stopped

LangFlow Embedded

You can embedd langflow in your websites with few lines of JS and this project: https://github.com/langflow-ai/langflow-embedded-chat

These other 2 are used for:

1. MLflow is an open-source platform to manage the complete machine learning lifecycle, including experiment tracking, model versioning, deployment, and providing integrated observability features, especially for LLM-powered applications.

2. Langfuse is an open-source observability platform specifically designed for LLM applications, offering detailed tracing, monitoring, and analytics to help developers understand, debug, and improve their language model-based systems.

LangFuse is kind of langsmith, but MIT and selfhostable

Once you create a local account, access the UI:

Rivet AI

DifyAI

Is Dify the easiest way to build AI Applications?

Conclusion

I have been tinkering with docker and some of these flows/bot builders:

See also…

How about embedding a chatbot within your website?

A branded AI-powered website chatbot

I show recently how to do a Streamlit Web App Chat bot.

But that might cause friction on users having to leave your home/services page and go to the streamlit domain/subdomain

You could do something similar, but inside your web, with:

ChatWoot

Docker Config 🐋 ↗

N8n Docker Config

Docker Config 🐋 ↗

ChatWoot

1. Chatwoot

N8N

2. What is n8n?

• https://github.com/n8n-io/n8n

• https://docs.n8n.io/hosting/

  • https://docs.n8n.io/hosting/starter-kits/ai-starter-kit/
  • https://docs.n8n.io/choose-n8n/#platforms

• Low-Code/No-Code, but with Code Flexibility: n8n is often described as a “low-code” platform. This means it provides a visual, drag-and-drop interface for building workflows without needing to write code for basic tasks. However, it also allows users (especially developers) to insert custom JavaScript or Python code within nodes for more complex logic and transformations.

• Node-Based Architecture: Workflows in n8n are built by connecting “nodes.” Each node represents a specific action, trigger, or data transformation (e.g., an HTTP request, a data manipulation, sending an email, or interacting with an AI model).

• Self-Hostable and Open-Source: A significant advantage of n8n is that it’s open-source and can be self-hosted. This gives you full control over your data, privacy, and infrastructure, which is a major differentiator from many other cloud-only automation platforms. They also offer a cloud-hosted version for convenience.

• Extensive Integrations: n8n boasts hundreds of pre-built integrations with popular applications, databases, and cloud services (e.g., Slack, GitHub, Google Sheets, Salesforce, HubSpot, OpenAI). If an integration doesn’t exist, you can use the generic HTTP Request node or create custom nodes to connect to almost any API.

• AI and LLM Capabilities: n8n is well-positioned for the “AI age.” It has native support for integrating with AI services like OpenAI’s GPT models, Claude, and Hugging Face. This allows you to build sophisticated AI-powered workflows for tasks like summarizing content, generating replies, classifying data, and creating AI agents.

• Use Cases: n8n is used for a wide range of automation, including:

  • Automating repetitive tasks across various tools.
  • Building complex data pipelines (ETL).
  • Creating custom integrations between applications.
  • Marketing automation (lead capture, email campaigns).
  • CRM integration and data synchronization.
  • IT Operations (onboarding, security incident enrichment).
  • Building AI agents.
  • Customer support automation.

Important Considerations when sharing/importing JSON:

• Credentials: Workflow JSON files include credential names and IDs. While IDs aren’t sensitive, the names could be. Crucially, they generally do NOT include the actual sensitive credential data (like API keys). When you import a workflow, you’ll need to manually re-link or create new credentials in your n8n instance for the workflow to function.

• Webhook URLs: If a workflow uses a Webhook trigger, the URL will be specific to the original n8n instance. When you import it, you’ll likely need to generate a new Webhook URL for your instance and update any external services configured to call that webhook.

• Node Versions: Occasionally, if you import a workflow built with a much older or newer version of n8n, you might encounter minor compatibility issues or warnings about outdated nodes. These are usually easy to resolve by updating the nodes within the UI.

In summary, n8n is a highly capable and developer-friendly automation tool, and its reliance on JSON for workflow definition makes it incredibly flexible for sharing, managing, and integrating with your existing development workflows.

We can embed n8n Chats into our sites thanks to: https://www.npmjs.com/package/@n8n/chat

npm install @n8n/chat #with this approach, you can customize how it looks, its called the import embed

But you can also use the script provided at the same npmjs link to have the default one, working out of the box, when placed at your website’s </head>:

<link href="https://cdn.jsdelivr.net/npm/@n8n/chat/dist/style.css" rel="stylesheet" />
<script type="module">
	import { createChat } from 'https://cdn.jsdelivr.net/npm/@n8n/chat/dist/chat.bundle.es.js';

	createChat({
		webhookUrl: 'YOUR_PRODUCTION_WEBHOOK_URL'
	});
</script>

This is called the CDN embed

1. Select add new trigger: Add on Chat message
2. Make Chat publically available & Select Mode as: embedded Chat

3. Now, get the YOUR_PRODUCTION_WEBHOOK_URL to feed it to the script
4. Paste the n8n workflow url into your website’s code. it will be something like: http://localhost:5678/webhook/some-random-uuid/chat`

You should soo this in the bottom right of your page:

And now…how to actually send messages and do stuff with n8n via your website?

5. Save the workflow and mark it as active via n8n UI

There are many more possibilities by playing with: https://docs.n8n.io/api/#learn-about-rest-apis

• https://docs.n8n.io/integrations/community-nodes/installation/
  • https://docs.n8n.io/integrations/builtin/app-nodes/n8n-nodes-base.mailerlite/
  • https://docs.n8n.io/integrations/builtin/app-nodes/n8n-nodes-base.googledrive/
  • https://docs.n8n.io/integrations/builtin/app-nodes/n8n-nodes-base.stripe/
  • https://docs.n8n.io/integrations/builtin/app-nodes/n8n-nodes-base.telegram/
  • https://docs.n8n.io/integrations/builtin/app-nodes/n8n-nodes-base.youtube/
  • https://docs.n8n.io/integrations/builtin/app-nodes/n8n-nodes-base.invoiceninja/
• https://docs.n8n.io/integrations/builtin/node-types/

• https://github.com/n8n-io/self-hosted-ai-starter-kit

The Self-hosted AI Starter Kit is an open-source template that quickly sets up a local AI environment. Curated by n8n, it provides essential tools for creating secure, self-hosted AI workflows.

• You can build cool things with these:
  • https://formbricks.com/docs/xm-and-surveys/core-features/integrations/n8n
  • Or…embedable chat widget - https://www.npmjs.com/package/@n8n/chat
    • https://docs.n8n.io/integrations/builtin/core-nodes/n8n-nodes-langchain.chattrigger/#make-chat-publicly-available

You can get ideas on what to build with n8n: https://n8n.io/workflows/

Like this one: https://n8n.io/workflows/4827-ai-powered-whatsapp-chatbot-for-text-voice-images-and-pdf-with-rag/

But for doing a quick N8N chatbot, like I did here, you are good to go with the OSS License

    environment:
      - N8N_CHAT_ENABLED=true
      - N8N_CHAT_CORS_ORIGINS=* #for local
      - N8N_CHAT_CORS_ORIGINS=https://your-website.com

So that this will become available in your n8n instance: http://192.168.1.11:5678/chat

Then, you can add the script to your website’s </head> as done here

• https://n8n.io/workflows/1954-ai-agent-chat/

TG Bots

3. Telegram…bots?

They can be used via n8n: https://docs.n8n.io/integrations/builtin/app-nodes/n8n-nodes-base.telegram/

But also code driven, if you dare.

Sometime ago I was doing some tests with tg bots, but that was all before the LLM boom.

And definitely there are interesting bots outthere, like the bibliotecasecreta.nl:

• https://core.telegram.org/bots/api
  • https://pypi.org/project/python-telegram-bot/
  • https://github.com/python-telegram-bot/python-telegram-bot
  • https://docs.python-telegram-bot.org/en/stable/telegram.ext.application.html

pip install python-telegram-bot==22.0

Its all done with Python and the following project structure:

With a .env file like this one:

BOT_TOKEN=YOUR_BOT_TOKEN_HERE
CHAT_ID=YOUR_TELEGRAM_CHAT_ID_HERE
OPENAI_API_KEY=YOUR_OPENAI_API_KEY_HERE

With the API token, you can use this token to access the HTTP API for your new bot.

Important Considerations for Bots in Groups:

• Privacy Settings: By default, bots in groups can only see messages that directly mention them (e.g., @YourBotName some command). If you want your bot to be able to read all messages in the group, you need to adjust its privacy settings via BotFather.
  • Talk to BotFather in Telegram.
  • Send the command /setprivacy.
  • Choose your bot from the list and see which one it has.
  • If its “Disable” your bot can read all messages. Be mindful of the implications of this setting.
• Permissions: As the group creator or an admin, you can manage the permissions of both regular users and bots within the group (e.g., who can send messages, add members, etc.).

For me, using /mybots allow me to change it.

• To obtain your chat ID on Telegram, you can follow these steps:
  • Open the Telegram app on your mobile device or use the Telegram web version on your computer.
  • Start a chat with the “userinfobot.” You can search for it by its username: @userinfobot.
  • Once you’ve started a chat with @userinfobot, send any message to it.
  • The @userinfobot will reply with information about your chat, including your chat ID. The chat ID will be a numerical value.
  • Copy the chat ID provided by @userinfobot. This is the ID you will use in your Python code as the chat_id variable.

And if you are already getting ideas… https://core.telegram.org/bots/webapps

FAQ

F/OSS No Code Data Tools

• Node Red
• Apache Airflow

BaseRow

• The open platform to create scalable databases and applications—without coding - https://baserow.io/

https://gitlab.com/baserow/baserow https://gitlab.com/baserow/baserow/-/blob/develop/LICENSE

Can be SelfHosted or use: https://baserow.io/database

Baserow is an open source no-code database and Airtable alternative.

Create your own database without technical experience. Our user friendly no-code tool gives you the powers of a developer without leaving your browser.

• Focus: Open-source no-code database tool with a user-friendly interface resembling spreadsheets but with relational database capabilities.
• Strengths: Offers a visual interface for building tables, defining relationships, and managing data. Supports various data types, including files, formulas, and automations. Integrates with popular tools like Zapier and webhooks.
• Considerations: While feature-rich, it might not be as powerful as some options for complex relational modeling.

Others

• API Table - https://github.com/apitable/apitable

  • https://github.com/apitable/apitable?tab=AGPL-3.0-1-ov-file#readme
  • https://github.com/apitable/apitable#installation

• Rowy: https://github.com/rowyio/rowy?tab=readme-ov-file

Focus: Low-code backend platform for building internal tools and automating workflows.

Strengths: Flexible and customizable, allowing you to build custom interfaces, integrations, and automations. Supports a variety of data sources and connects to external APIs.

Considerations: Requires more development knowledge compared to other options. Might have a steeper learning curve for non-programmers.

4. Appsmith: https://www.appsmith.com/integration https://github.com/appsmithorg/appsmith

Focus: Open-source low-code platform for building internal tools and custom applications.

Strengths: Visually build user interfaces using drag-and-drop functionality and connect to various data sources. Offers features like user authentication, form handling, and data integrations.

Considerations: While offering a visual interface, it might require some coding knowledge for complex functionalities.

Platform to build admin panels, internal tools, and dashboards. Integrates with 25+ databases and any API.

• https://docs.appsmith.com/getting-started/setup/installation-guides/docker?utm_source=github&utm_medium=organic&utm_campaign=readme

services:
   appsmith:
     image: index.docker.io/appsmith/appsmith-ee
     container_name: appsmith
     ports:
         - "80:80"
         - "443:443"
     volumes:
         - ./stacks:/appsmith-stacks
     restart: unless-stopped

5. Budibase:

Focus: Open-source self-hosted alternative for building internal tools and managing data.

Strengths: Offers a modular architecture with components for building databases, forms, dashboards, and workflows. Supports user management, role-based access control, and data filtering.

Considerations: Relatively new project compared to some alternatives. Feature set might be under development.

• https://github.com/Budibase/budibase
• https://github.com/Budibase/budibase?tab=License-1-ov-file#readme

Low code platform for building business apps and workflows in minutes. Supports PostgreSQL, MySQL, MSSQL, MongoDB, Rest API, Docker, K8s, and more 🚀

• https://github.com/RasaHQ/rasa
  • https://github.com/RasaHQ/rasa?tab=Apache-2.0-1-ov-file#readme

Rasa Open Source is the most popular open source framework for building chat and voice-based AI assistants.

💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants