Super Nerdy AI Research Automation

I actually created this spreadsheet that pulls in new research papers and then it gives me a one sentence summary of them and then gives it a tag of like this is related to LLMs, image generation, video generation, 3D object creation. And then you can actually see the title of the paper here. So I have this feedly account here and it pulls in all of the latest research from hugging face automatically. And then with my make.com account it's watching that feedly account for new posts that come into it. And whenever it finds a new post it uses a tool called scraping bee and it scrapes the abstract here. It pulls whatever it scraped into Claude and you can see, see here I gave it the prompt. Below is a scraped page with the abstract from a research paper based on the title and abstract. Please write a one sentence headline that says exactly what the research does in simple terms with no technical jargon. It should read like a news headline that's sharing a breaking announcement about what this research is and why people should care. It pulls in the title and then content of site. This is pulling in anything scraping be just scraped. And then it pulls the response from that into another claw chat here and this is the one that tells it to give it a tag. And then I use this text parser here. All this is doing is removing some extra stuff that's in the title and then it dumps it into Google sheets for, for me you can see the date, it pulls in today's date, the research title, the research URL. We've got our description in our category and you can see that it's just pulling those in from what it generated from Claude. And then I have it wait one minute. So it doesn't just rapidly fire these off one after another because Claude will get bogged down and say too many requests.