This is a category for showing off the projects that you’ve built with n8n.
Everyone is invited to ask questions to the authors of the topics about their projects.
Examples of posts that can be shared here include processes that you have automated using n8n workflows and apps or products that you’ve built that are powered by n8n.
Note: A better place to showcase single workflow projects is the workflows page.
I am a new n8n user with a strong eagerness to learn. I am an IT graduate, and while professional responsibilities have kept me working with more traditional technologies, I am now intentionally refocusing my efforts on AI and automation to upgrade my skill set and pursue new career opportunities. I look forward to learning from this community and contributing as I continue to grow professionally. This is what im tryin to build…
Chat-Based Order Processing Workflow (Concept Overview)
I am designing a chat-based order automation workflow using n8n and AI with the following objectives:
Customer Interaction Channel
Customers initiate orders by sending messages to my Facebook Page Messenger.
AI-Powered Message Processing
Incoming messages are received by the workflow and processed by an AI component that:
Interprets unstructured chat messages
Extracts structured order information
Multilingual Support (Waray-Waray to English)
Since most customers will not use English, the AI will:
Detect messages written in Waray-Waray (Philippine dialect)
Translate the content into English prior to processing
Use the translated text as the basis for order understanding and extraction
Order Data Extraction Requirements
The AI should extract and normalize the following details:
Ordered Items
Item name
Quantity
Order Source
Vendor or store name
Support for multiple items from multiple sources within a single message
Delivery Location
Barangay or specified delivery area
Example Input (Customer Message):
“Chicken 1pc from Jollibee and burger plain 2pcs from McDonald’s, deliver to Barangay E.”
Workflow Goal (Initial Phase)
The primary goal of this first iteration is to reliably transform free-text chat messages into clean, structured order data that can later be:
Stored in a database or Google Sheets
Used for order confirmation messages
Integrated with downstream fulfillment or delivery workflows
I have built so many project with the help of n8n like Voice Assistant Agent that controls my social Media and Google account . If anyone wants to work with me . Contact with the page https://portfolio-ai-assistant-of-malik.vercel.app/
Building a Local-First Automation Portfolio with n8n
Over the last weeks I’ve been building a local-first automation environment around n8n running on Apple Silicon.
The goal was not to build isolated workflows, but to explore how far a self-hosted setup could go when treated as an internal automation platform rather than a simple workflow editor.
The stack currently runs locally using:
n8n
PostgreSQL
Docker + Colima
External-disk backed persistent storage
Automated backups and validation scripts
Secret scanning and release packaging
The project currently consists of three workflow case studies:
1. Job Application Intelligence Pipeline
A workflow for importing, validating, enriching and persisting job application data into PostgreSQL with import auditing and reporting.
2. Fleet Operations Automation System
A workflow exploring operational visibility, vehicle tracking and fleet-related processes inspired by real operational environments.
3. Local AI Knowledge Processing Workflow
A local-first document ingestion and knowledge processing pipeline focused on privacy and self-hosted operation.
Some of the more interesting challenges were not the workflows themselves but the surrounding operational details:
PostgreSQL persistence verification
File access restrictions
Container path visibility
Import/export packaging
Backup validation
Secret sanitization for public releases
Reproducible local deployments
One issue that took some investigation was file access restrictions inside n8n containers and ensuring workflows could only access explicitly allowed paths through N8N_RESTRICT_FILE_ACCESS_TO while still maintaining a usable import process.
The entire project is now published publicly:
GitHub:
Live presentation:
I also recently joined the community under the username kostasuser01gr and started participating in discussions around PostgreSQL, AI-assisted queries and workflow troubleshooting.
I’d be interested to hear how others approach local-first deployments, workflow packaging and release management around n8n projects.