Ai Automation Complete Course - n8n, Zapier Automation Bangla Tutorial

Technology concepts & product features covered

1) AI Automation Master Class (overall course structure)

• Presented as a step-by-step “master class” for learning automation from fundamentals to intermediate projects, organized into chapters.
• Teaches how to build AI agents and automated workflows, then monetize them (domain/hosting + earning via automation).
• Mentions example learning paths/projects such as:
  • Instagram lead generation agent
  • AI-powered roadmap/search agent
  • A trading/content/prediction-themed agent (project mention)
  • Newsletter automation agent (generate/optimize/publish/send dynamically)
  • Final earning/micro-product approach (monetization + hosting)

2) Core AI Concepts: What is an AI Agent?

• Defines an AI agent as software that can:
  • perceive an environment
  • make decisions
  • act independently to reach a goal
• Includes a real-world analogy: a digital assistant managing tasks like email/calendar.

• Explains types of AI agents:

  • Reactive agents: respond directly to input Examples: Google Assistant, Siri

  • Model-based agents: use past knowledge/model Example: a robot vacuum using room maps

  • Goal-based agents: optimize for a specific goal Example: Tesla self-driving route planning

  • Utility-based agents: optimize user/platform experience Examples: YouTube/Netflix recommendations

  • Learning agents: continually improve Examples: ChatGPT/Gemini-style learning from feedback

3) Automation vs AI Automation (key distinction)

• Traditional automation: rule-based, fixed instructions Example: send the same email at 9 AM

• AI automation / AI agents: can learn/adapt/decide using data and context Example: personalized emails based on user behavior

• Typical skill-set contrast:

  • Automation: focus on workflow setup
  • AI: involves ML/data science/deep learning/data processing

4) Tool stack introduced: Zapier, n8n (NET), LangChain, Langflow

Zapier (JPR in subtitles)

• Marketed as a no-code automation powerhouse for connecting “50M+” tools/apps (subtitle claims “5000 tools/apps”) via Zaps.
• Core workflow idea:
  • A Zap = Trigger + Action
  • Supports multi-step workflows
• Demonstrated use case:
  • Google Forms → store data → Gmail sends personalized replies without coding
• Subtitles emphasize:
  • trigger types like new/updated form responses
  • dynamic field mapping (email/name from the form)

n8n (“NET” in subtitles)

• Described as open-source workflow automation, featuring:
  • self-hosting (local via Docker/localhost; also hosting on a server)
  • more flexibility/control than Zapier
  • drag-and-drop plus custom code
  • integration with APIs
• Core UI concepts:
  • Workspace
  • Project
  • Workflow
  • Nodes
  • Execution (runs the workflow)
• Node structure explained:
  • A node is one step/action in a workflow
  • Node types include trigger nodes, action nodes, transformation/logic control nodes, external service nodes
• Deployment modes discussed:
  • cloud, local, and self-hosting

LangChain (“Langchain/Langchen” in subtitles)

• Presented as a backend framework for AI agent pipelines:
  • memory, reasoning, decision-making via chains/models

Langflow (“Langflo/Langflow” in subtitles)

• Presented as a visual UI (front-end) for building agent pipelines.
• Key split:
  • LangChain = backend logic
  • Langflow = drag-and-drop interface controlling how the agent processes/thinks/responds

5) Practical n8n tutorial: Build a basic Zap-like email automation

• Step-by-step n8n workflow analogous to “Google Forms → Gmail personalized email.”
• Workflow includes:
  • Create a Google Form
  • Trigger: “new response”
  • Action: send email via Gmail
  • Dynamic field mapping:
    • “To” uses submitted form email
    • body/subject/name uses submitted name for personalization
• Includes testing/debugging notes:
  • check wait time
  • check spam folder if email doesn’t arrive

6) n8n practical project: AI “Joke Generator Agent” using OpenAI + form input

• Form-based workflow:
  1. User inputs name + joke topic
  2. n8n executes
  3. Calls an LLM via OpenAI tool node (assistant creation + message prompting)
  4. Returns/generated joke text
  5. Displays the output on a result page
• Highlights:
  • system prompt vs user prompt
  • selecting an OpenAI model (e.g., GPT-4.0/GPT-4o-mini mentioned)
  • generating output in English, or optionally Bengali (prompt instruction)

7) Lead Generation Agent project: Instagram leads → save to Google Sheets

• Described as a 4-step Instagram lead generation agent (also mentions adapting to LinkedIn/Facebook/Twitter).
• Key flow:
  • Create a form to collect:
    • keyword (e.g., “automation expert”)
    • location (e.g., Bangladesh)
  • Use a scraping platform (Apify referenced as “Efi/efifi”) to scrape search results
  • Use an HTTP request node to call the Apify API:
    • query variables derived from the form (keyword/location)
  • Use a code node to clean messy Apify JSON into structured fields:
    • name, profile URL/handle, followers, description, etc.
  • Append results into Google Sheets:
    • append row action with field mapping
• Mentions adjusting results per page (e.g., 50/100 depending on plan)

8) Advanced n8n + Airtable project: save form data and use conditions (switch node)

• Builds a project to:
  • collect user inputs via a form (name/email/phone/summary; later adds profession)
  • save into Airtable using Airtable API credentials
• Covers data formats:
  • JSON vs Schema
• Connection setup:
  • Airtable API using Personal Access Token
  • required scopes mentioned: read/write + base/schema access
• Debugging/data typing issues:
  • Airtable phone field expects formats including country code
  • solution: store phone as text instead of numeric phone type
• Utility logic:
  • Switch node for routing:
    • if profession = programmer → programmer roadmap path
    • if designer → designer roadmap path
    • if AI Automation Expert → AI automation roadmap path

9) AI roadmap generator agent using Gemini + n8n AI agent node

• Connects an LLM (Gemini / “Jiminy” in subtitles) to generate a personalized roadmap using:
  • name, email
  • profession selection
  • user prompt template input
• Uses AI agent functionality inside n8n:
  • chat model integration + credentials setup (Gemini API key)
  • a system prompt defining the role (e.g., mentor/tutor creating step-by-step roadmap)
• Saves output back to Airtable:
  • create/update Airtable record
  • update only the roadmap field with AI response
• Mentions prompt constraints:
  • keep roadmap roughly under ~1000 words
  • output in Bangla when requested

10) Utility node / helper node concept

• “Utility node” described as a helper for:
  • manipulating data between steps
  • handling control flow when normal node connections aren’t enough
• Example: conditional routing / multi-path logic (ties into Switch node usage)

---

Reviews / guides / tutorials explicitly present

• Beginner-to-intermediate “master class” tutorial format with multiple projects.
• Detailed guides for:
  • Zapier-style Trigger→Action workflow
  • Google Forms integration + Gmail personalized email mapping
  • n8n interface concepts (workspace/project/workflow/nodes/executions)
  • Building an OpenAI/Gemini-based AI agent with system/user prompts
  • Lead generation automation using Apify HTTP API + code node cleanup + Google Sheets append
  • Airtable integration using API token + schema/table mapping + debugging phone number formats
  • Conditional routing using Switch node
  • LLM roadmap generation and saving results to Airtable

---

Main speakers / sources (as inferred from subtitles)

• Instructor: “Aviwan” (repeated address: “Hey Aviwan…”)
• Course/Channel host: “Hublu Programmer” (referenced as the instructor identity/channel name)
• Automation platforms and APIs referenced as tools/sources:
  • Zapier
  • n8n
  • LangChain / Langflow
  • OpenAI (and “ChatGPT”-style prompting)
  • Gemini (Google)
  • Apify (scraping provider)
  • Google Forms / Gmail / Google Sheets
  • Airtable