Launching today
ScrollWise AI

ScrollWise AI

Your research powered by the world's most powerful models

4 followers

Visit website
Dump articles, tweets, PDFs, and notes into topic collections. Search semantically to find that hidden truth nuke. Ask questions to the world's most powerful models, and get answers grounded in YOUR sources. No more vague responses. Power your research with relevant, timely content and the world's premier AI models.
ScrollWise AI gallery image
ScrollWise AI gallery image
Free Options
Launch tags:Notes•Social Media•Artificial Intelligence
Launch Team / Built With
AssemblyAI
AssemblyAI
Build voice AI apps with a single API
Promoted

What do you think? …

Mitchell Edwards
GrabbrApp
Maker
📌
Hey all! My name is Mitch, and I have research ADHD. I have normal ADHD as well, but research ADHD to me means hoarding a ton of resources related to a hundred different topics and... never being able to find or use any of them. Over the years, I've done research on cyber security, political extremism, blacksmithing, software development, business, Mandarin Chinese and more. My problem has been that my data is everywhere: 30 notes apps (I use and love Obsidian, btw), 1,000,000 bookmarks and To Read tabs, and a thousand tweets in my bookmarks. When I try to use LLM's, the answers they give are vague, outdated and sometimes wrong. The models are extremely powerful, but they don't always have access to up-to-date, relevant information for what I'm looking into. ScrollWise is the answer. In its current form, you can add Tweets, Substack articles and raw text "Resources" to various "Topics" that you're interested in. Those resources are consumed and analyzed, letting you run semantic search across all of your favorite online content. Then, when you query your favorite models using the ScrollWise Chat, relevant content is injected into your conversation, making the models smarter and giving them access to the up-to-date, specific context they need to give you a useful answer. Coming soon (like, this week or next): - YouTube transcript extraction (pulling text from a YouTube video) - PDF parsing (pulling text from PDF files uploaded to the platform) - API keys (for the nerds like me that want to automate the whole process) - Multi-model support (for the people who, for some reason, don't just use Claude for everything) - Personal notes (for adding context to the material you're gathering) Coming later (within the next few months): - A browser plugin (push content to your Topics collections without leaving the page) - Public notes (publish your research, whether it's for clout or to spread the word) - Enhanced context engine (I'll be working hard to make sure that the context added to your chat conversations is as useful as possible)