Hi everyone! 👋 Just shared a quick 15-min tutorial showing how to build a smart AI agent in N8N that can chat directly with PostgreSQL/Supabase databases - no vector stores needed! It's especially useful if you're working with dynamic data like pricing, services, or product details that change frequently. Would love to hear your thoughts and experiences if you try this approach!

Hi guys! I just finished building a simple tool for my client's dental clinic - we automated call analysis with OpenAI and AssemblyAI. We parse call with AssemblyAI and detect bunch of metrics including upsell opportunities. If it exists, we send a notification to another manager who is responsible for upsells. Feel free to copy it and offer your own customers - every business with sales calls needs this. It's easy to sell this bcs, there are bunch of solutions on the market, but most of them for enterprise and annual cost is much more than payment to you for integration + OpenAI/AssemblyAI expenses.

Hello everyone! I made a tutorial with the fastest way to self host your own N8N instance with DigitalOcean. It's not a simple version of N8N - there are custom JS libraries (OpenAI, Supabase and etc) and PostgreSQL DB instead of default MySQL. I also simplified the whole process into 5 commands with special scripts to make it possible to selfhost in 10 minutes. Why n8n and this setup? 1. It's cheaper and more flexible than Zapier and Make (telling you as ex-Make QA engineer 😄) 2. You can build AI agents and use JS/Python code with custom libraries. I have added instructions for easy setup of JS libraries. 3. The default MySQL DB that holds execution history and N8N data is weak, and scaling scenarios with DB can soon lead to speed and memory issues. So I've added PostgreSQL DB instead, which is automatically installed on DigitalOcean.