n8n Matrix Display

In the 1987 film Wall Street, the stock quotes flashed across a glowing dot-matrix display mounted on the wall — a mesmerizing mix of technology and style.

Over the years, I’ve built several scrolling displays using all sorts of hardware and code. During the.com boom, I made a primitive version using a locally produced Microled panel that came with a very basic software for static messages. I hacked together a macro based method to inject live sales updates and early Internet news. Later came micro-displays built from 16x2 LCD screens, and one huge version made from multiple HDMI monitors.

Now, I’m returning to that tradition — but this time, powered by the automation muscle of n8n.

What is n8n?

n8n is an open-source workflow automation platform that connects APIs, databases, and online services. How does that help a scrolling display? It lets you create a dynamic feed built from multiple sources, with a touch of AI intelligence in the mix.

In my setup, the display shows the current date and time, weather, available parking spaces (via Gemini Flash multimodal LLM), unread Gmail messages, Nasdaq quotes, and tech news. Of course, there’s no limit to what you can build — whether pulling from your own data or mixing public sources.

Hardware

The hardware is simple, affordable, and easy to source:

2 x WS2812B 16x16

1 x Beetle ESP32 C3

1 x Power Supply 5V 3A

Plus a few 3D-printed parts.

3D-Printed Parts

I designed a frame for the LED panels, connectors, and an ESP32 enclosure using Fusion 360, and printed them on a Bambu Lab A1 Mini with PLA. The files are available on Cults3D. Use six 3 mm screws to connect both panels and attach the back cover.

Circuit

The circuit couldn’t be simpler: link the two panels using the connector included with the WS2812B. Power the first panel from the 5V external supply and connect it to the ESP32 through a single data pin.

Workflow

You can start using n8n with a free 15-day trial at n8n.cloud, or by hosting a pre-installed version on Railway, Hostgator or any other service. Self-hosting works too.

The ESP32 calls a Webhook node in n8n, which replies with the text to display. Between these two nodes lies the logic — a small AI agent with access to several tools. Adding new info to the display is as simple as adding a new tool and describing how to use it in plain language. n8n offers tool nodes for virtually every use case, and you can always build your own.

If you want to use my example workflow, just import it, set up your credentials for Gmail, Gemini, Market Stack, etc., and change the agent’s language prompts. You’ll also need to set the URLs for the parking-lot camera images (more about this below)

ESP32 Code

Before uploading the sketch to your ESP32, configure the webhook URL — you can find it by opening the starting Webhook node in your workflow. Also set your Wi-Fi SSID and password.

Select ESP32C3 Dev Module in the IDE and use the specified settings.

AI-Powered Parking Detection

Until recently, detecting free parking spots through computer vision required training a custom machine-learning model. Now, multimodal LLMs make this process much easier. Mount one or more cameras pointing at your parking area or street. At regular intervals, capture images and upload them to a web server. This can be done with ESP32-Cams, Raspberry Pi, or even an old laptop and webcam using Python.

Once the image is uploaded, n8n’s tool sends it to Gemini 2.5 Flash with a prompt asking whether the model “sees” a free parking spot.

The Python code for capturing and uploading images is available on GitHub — just configure your webcam ID and FTP credentials.

Retro Gaming Animations

To make the display more fun, I also added some retro game animations: Pong, Pac-Man, and Space Invaders.

Final Notes

Whether you want a live-updating stock ticker straight out of Wall Street, public transit info, local weather, work-related alerts, AI-interpreted database reports, or a digest of the latest maker projects from Hackster — combining these affordable LED panels with n8n lets you design your own custom data feed without wrestling with dozens of APIs. The results can be surprisingly powerful.