I built a pretty epic Lovable + n8n workflow, wanted to share it with you guys. This one uses lovable with webhooks from n8n to run the backend. It... - Uses AI to give me "high intent" questions that people are asking, - Let's me record a video to answer the question - Automatically transcribes and adds subtitles - Draftgs a linkedin post - Let's me edit the linkedin draft - Let's me post the linkedin post Took a few weeks on and off to figure out, it's also a bit complex, but I like the workflow a lot. - Lovable for the frontend - Railway with a python server using WhisperX + ffmpeg for the subtitles (largely vibe coded) - n8n to hold everything together. Just posting so you guys can see something practical; my client is super excited for this automation, it enables better videos + only takes 5 minutes / day for the whole process. P.S. I'll post something for the agency in a box challenge soon too.

Did my ChatGPT Year Review and this is right on par for how my year went. I spent a LOT of time with AI this year and it’s got me really curious about what this year will bring for me… especially in this new era of vibe coding and wanting to get an app published to the App Store this year. Looking forward to 2026. Cheers 🥂

Just want to share I use this app WhisprFlow (https://wisprflow.ai/) to do voice to speech transcription. I fill out a lot of forms, and this speeds up the process by like 5x. Thought I would share, especially when prompting AI this helps a lot because you're typing a lot. It also makes you sound more conversational when sending email or stuff like that. Enjoy!

Hey AI-ers, And lovable people! Here's my workflow with Lovable, that works best for me to build! 1) Refine my idea with Claude. "I'm thinking of building x, can you ask me some questions and make a PRD for this to give to Lovable" 2) Give PRD to lovable (and maybe some others, are we allowed to do that here?) 3) Once I'm generally happy, I'll push to github 4) Download my github repo in VS Code / Cursor, then continue with copilot / claude code. 5) Usually I need to tell the AI to clean up the code and organize it with DRY / Functional programming, and I built out templated elements so that it's easier to build out as I go. 6) Publish to Vercel -> I find this to be the easiest, and I have full control. It's maybe not as nice as doing all in Lovable, but it gives me a lot of control. When I work with Lovable I find as time increases the likelihood of me going in circles also increases. If anything is unclear, let me know. Hope this helps you 🙏

Was tired of 1x nano banana images. Then I had to manually convert png to webp for my website projects. So I built an n8n workflow, then used MCP to build a frontend in lovable, and now I can generate 2 images instead of 1 AND I can download them as webp. The webp part took me awhile, first I tried to do it with n8n, but needed extra APIs, turns out lovable could do it easier for me. Since google gives me $300 free credits, it also free-er than nano banana haha. Just sharing as an idea for you guys what's possible. Lovable is a great frontend, n8n is a great backend.