Customer Feedback Sentiment Analysis (n8n + Gemini)
Built an AI-powered feedback system that reads incoming customer submissions, classifies the sentiment, and automatically responds and logs the result, with no manual review needed. When a form is submitted, an AI agent analyzes the feedback and routes it down one of two paths based on sentiment. Positive feedback is logged to Google Sheets and answered with an AI-generated thank-you reply sent via Gmail. Negative feedback goes through an additional Suggestion AI Agent that extracts what needs fixing internally before the complaint is logged separately and an apology email is drafted and sent to the customer.
Tools used:n8n, Google Gemini, Gmail, Google Sheets, conditional branching logic.
RAG-Powered AI Knowledge Agent (n8n + Supabase)
Built an AI agent that answers questions using retrieval augmented generation, pulling real-time context from a personal knowledge base rather than relying on generic AI responses. Documents are automatically ingested from Google Drive, chunked, converted into embeddings using OpenAI, and stored in a Supabase vector database. When a user asks a question, the AI agent searches the vector store for the most relevant information and generates an accurate, grounded response, while Postgres-based chat memory maintains context across the conversation
Tools used: n8n, OpenAI (Chat + Embeddings), Supabase Vector Store, Google Drive, Postgres Chat Memory.
RAG-Powered AI Knowledge Agent (n8n + Supabase)
Built an AI agent that answers questions using retrieval augmented generation, pulling real-time context from a personal knowledge base rather than relying on generic AI responses. Documents are automatically ingested from Google Drive, chunked, converted into embeddings using OpenAI, and stored in a Supabase vector database. When a user asks a question, the AI agent searches the vector store for the most relevant information and generates an accurate, grounded response, while Postgres-based chat memory maintains context across the conversation.
TOOL USED:
n8n, OpenAI (Chat + Embeddings), Supabase Vector Store, Google Drive, Postgres Chat Memory.
AI Email Triage & Auto-Responder
A fully automated support inbox pipeline that reads every incoming email, classifies it by category and urgency using an AI agent, and routes it accordingly. Routine emails are answered instantly with an AI-drafted reply sent straight back to the customer. High-urgency emails skip the queue entirely, the internal team is alerted immediately via Telegram, and a draft reply is prepared in Gmail so a human can step in fast, without starting from a blank page.
WHAT IT DOES:
Cuts response time on routine tickets from hours to seconds.
Urgent issues are never buried under low-priority ones.
Human stays in control on sensitive cases, automation drafts, a person decides.
SYSTEM SUMMARY: SOCIAL MEDIA CONTENT CREATION AGENT
• WHAT IT DOES:
Uses autonomous AI agents to research a topic and generate ready-to-post content for LinkedIn, Facebook, and blog, delivered straight to Google Sheets.
• HOW IT WORKS:
TRIGGER: New row added to Google Sheets starts the workflow.
RESEARCH: Tavily web search gathers current, factual info on the topic.
GENERATE: Three AI agents write platform-specific content in parallel, LinkedIn, Facebook, and Blog.
DELIVER: Finished content is written back into the sheet, ready to publish.
• VALUE POINTERS:
✓ Cuts content creation time from hours to minutes.
✓ Platform-matched tone and structure, automatically.
✓ Fact-grounded via live web search.
From RSS Feeds to Inbox: An AI News Automation
A fully automated news pipeline that runs on a set schedule, pulling the latest stories from multiple RSS feeds and merging them into a single, unified stream. Rather than dumping raw headlines into an inbox, an AI summarization step reads through the articles and condenses them into a clean, digestible brief, capturing the key stories without the noise. The finished digest is then delivered automatically via email, giving you a ready-made update waiting in your inbox every day, with zero manual effort required from trigger to delivery.
LinkedIn Post Generator n8n AI Agent Workflow
An end-to-end automation that turns a form submission into a fully published LinkedIn post. A chain of AI agents (powered by OpenAI) generates the post description, image prompt, and title, grounded in real-time data via web search. The workflow then creates a matching image, uploads it to Google Drive, publishes the post, and sends a confirmation email via Gmail
Key features:
Multi-agent AI pipeline (content generation → image prompt → title generation)
Web search integration for accurate, current information
Automated image generation and cloud storage (Google Drive)
End-to-end publishing with email confirmation
Built entirely in n8n with OpenAI-powered agents
SYSTEM SUMMARY: AUTOMATED CRM EMAIL ASSISTANT
• WHAT IT DOES:
This workflow uses an autonomous AI agent to read customer data from a Google Sheet and automatically send personalized emails via Gmail based on simple chat commands.
• HOW IT WORKS:
1. TRIGGER: A user inputs a natural language command into the chat interface (e.g., "Send an email to Favour...").
2. BRAIN: The Groq Chat Model processes the request and calls the necessary tools sequentially to find missing details.
3. DATA LOOKUP: The "fetch_all_customer_rows" tool retrieves the customer rows, scanning for the matching name and email.
4. ACTION: The workflow dynamically maps the recipient details, crafts a custom body message, and executes the Gmail node.
• VALUE POINTERS:
✔ Eliminates manual data entry errors entirely.
✔ Speeds up standard customer notification delivery.
✔ Highly scalable layout easily adapted to platforms like HubSpot.
Automated Daily Inspiration Emailer
What It Does
An automated system that automatically writes and emails a personalized motivational message to every person on a subscriber list every morning.
How It Helps a Business
Instant Personalization: It greets and inspires each subscriber by their own name, making the brand feel personal and attentive.
Hands-Free Operation: It automatically pulls client details from a spreadsheet and sends the emails out without requiring any manual daily effort.
Mass Customization: Instead of sending the exact same generic blast to everyone, it uses AI to write a fresh, unique message tailored specifically to each individual reader.
This automation automatically saves, cleans, and tracks your data between different apps without any manual work
This weekly automation acts as a personal AI research assistant that finds new podcast topics, prevents repeat episodes, and sends a complete script straight to your inbox.