Susan Emmanuel-IkechukwuVerified StackedHQ member

I built an AI Office Assistant workflow in n8n that automatically reads incoming Gmail emails, classifies them using AI, and routes them to the correct category. The workflow starts with a Gmail Trigger, extracts the email subject and body, sends them to an AI Email Classifier, then merges the AI result with the original email before routing it through a Switch node. For invoice emails, the PDF attachment is automatically uploaded to Google Drive. Although the challenge mentioned using OpenAI, I used Groq instead. I ran into API quota limits on the OpenAI free tier, so I switched to Groq to k

I built an automated appointment reminder and SMS confirmation system for a medical clinic using n8n, Google Calendar, Twilio, Airtable, and a custom dashboard. The workflow checks upcoming appointments, verifies them a second time before sending reminders, prevents duplicate messages, skips cancelled or rescheduled appointments, and logs every outcome for a complete audit trail. I also created a dashboard where staff can manually trigger reminder checks, monitor delivery statistics, and view system activity in real time. The solution focuses on reliability, trust, and ensuring patients only r

For this challenge, I built Emma, an AI-powered virtual receptionist for Bright Smile Dental Clinic. Emma welcomes patients, answers enquiries, collects appointment details, qualifies leads, and automatically triggers a booking workflow. The system also sends confirmations, creates calendar appointments, updates records, and notifies the clinic team. I built it using n8n, AI Agents, Postgres memory, Google Calendar, Gmail, Telegram, and Google Sheets. One of the biggest improvements during this project was evolving Emma from a simple chatbot into a more reliable virtual receptionist with me

I built an AI CRM Lead Nurturing Assistant for a real estate business using n8n, HubSpot, Gmail, AI, and Airtable. The workflow starts automatically when a new contact is created in HubSpot through a webhook connected to a HubSpot Private App. It generates personalized follow-up emails, sends them automatically, and continues with scheduled follow-ups if there is no response. For the demo, I shortened the wait times to a few seconds, but in a real-world scenario they would be spaced over several days. I also built a Reply Monitor that detects incoming replies, stops the nurture sequence, updat

I built an AI-powered expense receipt processing workflow using n8n. Drivers simply send a receipt through WhatsApp, and the workflow automatically downloads it, stores it in Google Drive, extracts the receipt details using Google Gemini Vision, and saves the information to both Airtable and Google Sheets. It also checks for unsupported files, blurry receipts, multiple receipts in one image, and duplicate submissions before saving the data. If there's a problem, the driver receives a WhatsApp message explaining what needs to be corrected. If everything is valid, the receipt is processed succes

I built SolarFlow Lead Intelligence, an AI-powered lead qualification and sales intelligence system for a solar business. The system captures a new lead through a qualification form, uses AI to assess and score the lead, and then automatically sends the information to the CRM and other business tools. It also handles the follow-up and brings the important lead information back into the SolarFlow dashboard so the sales team can quickly see who needs attention. The goal was simple: less manual work, less jumping between platforms, and a clearer picture of which leads are worth following up with.

I built a voice AI receptionist for Bright Smile Dental Clinic that can answer patient calls, handle common questions, check appointment availability and book appointments directly into Google Calendar. Emma also logs calls and bookings in Airtable and Google Sheets, so the clinic has a record of what happened. I added separate handling for appointment changes, so rescheduling and cancellation requests are logged as their own records instead of overwriting the original appointment. This gives the clinic a clear audit trail of the original appointment, the requested change and whether human act

I built Smart Operations Router, a workflow that takes in business events and turns them into organized, trackable tasks, so every request actually lands with the right team the moment it comes in, not hours later. When an event comes in, it gets stored, cleaned up into a consistent format, and read by AI, which decides which department it belongs to, how urgent it is, and whether it needs someone's sign-off. From there, it's automatically routed to Sales, Support, Ops, or Reporting, a task gets created, and the right team gets notified in Slack. If approval is needed, Slack shows Approve/Dec