Kusa has become my go-to tool for analyzing my clients' markets (startups and solo entrepreneurs). What used to take me 2 weeks of niche research now only takes 2-4 days.
The service saves me hours: I can extract key insights from hour-long podcasts in just 20 minutes using prompts, and consolidate data from multiple sources into a single summary.
Thanks to Kusa, I've cut my pitch deck creation time from 3 weeks down to 1 - my clients love the speed.
I'd love font customization - that's the only feature that would make me ditch my other tools completely.
Perfect for anyone working with entrepreneurs who values their time.
Kuse
Hey PH fam!
Xiankun here, founder of Kuse.
Thanks Siqi for hunting us, and your incredible support since day one.
A year and a half ago, we started exploring a question: how can humans and AI collaborate? As someone who lives in Figma and whiteboards to organize my thoughts, our first attempt was simple: a canvas where you could place your files and your LLM convo cards. This allowed for a fluid way to connect, select, and combine information to interact with the models more freely.
After we launched, we were amazed to see users uploading a huge variety of file types, using the whiteboard as their primary space for reading and processing information. This pushed us to rapidly expand our file support and processing capabilities. Soon after, users began asking for a way to generate specific deliverables, like images, documents, and even web pages, based on their uploaded information. We quickly got to work to make that a reality.
Through this journey, our core mission became crystal clear: in an age of fragmented information, we want to provide a unified workspace. A place where you and your AI partner can process all your relevant information and create the exact outputs you need.
So with Kuse, our focus is to empower you to:
-> All Your Information: continuously expanding support for more file formats and data sources, bringing everything that matters to you into one place.
-> Intuitive UX: a unified, visual interface to combine and leverage your information (real-time team collab is also WIP!)
-> Smarter AI: to better understand your context, and current ideas and workflows
-> Wide Range of Outputs: docs, images, videos, web pages, and more, without switching between different tools
Even without spending a dime on marketing, we're humbled to have over 400 fast-moving teams and 200,000 top performers: educators, product managers, creatives, consultants, students, and writers, love Kuse.
I was fortunate to go through Y Combinator with a previous company (rct.ai, W19), where I learned one simple but profound lesson: "make something people want."
Your feedback, suggestions, and critiques are invaluable in helping us get closer to that goal.
Thank you so much for your support! ❤️
@xiankun Feels like a real thinking canvas. Do you support real-time collab, keyboard-first flow, and Notion/Drive import with live sync?
Kuse
@anwarlaksir Hi Anwar, real-time collab & sync is on the roadmap!
Toki: AI Reminder & Calendar
Mom Clock
Congrats on the launch, Xiankun!
How does Kuse’s AI determine the most relevant outputs—like documents or web pages—from a mix of uploaded files and conversation cards, and can users guide or customize that process?
Kuse
@justin2025 Thanks for the question, Justin! Kuse’s AI looks at the content across your files and conversation cards to find the most relevant information based on context and patterns. Users can definitely guide the process by organizing cards, tagging content, or highlighting what’s important, which helps the AI surface results that matter most. We’re excited to see how people customize it for their workflows!
Chance AI: Curiosity Lens
Really fascinating evolution from a simple canvas-LLM hybrid to this unified workspace concept. The YC background shows in how user behavior has guided the product development.
What really clicks here:
Visual canvas approach to organizing AI interactions feels much more natural than traditional chat interfaces
Expanding file format support based on actual user behavior was smart
Focus on generating specific deliverables (docs, images, web pages) directly from the canvas
That hint of "more canvas magic coming soon" is exciting - the platform already seems powerful for thinking and creating.
What stands out is how this tackles the fragmentation problem in knowledge work. Instead of adding another silo, it creates a space where different types of information can actually interact meaningfully.
Quick question: Have you seen any unexpected workflows emerge? Like combinations of file types or canvas arrangements that weren't part of the original vision but turned out to be really powerful?
The validation from educators, PMs, creatives, consultants makes sense - these are all roles that deal with connecting disparate information into coherent outputs. Really impressed by how this feels designed for actual thinking, not just another AI chat interface. The visual, flexible, outcome-driven approach makes a lot of sense.
Upvoted! Looking forward to seeing how the real-time collaboration features develop. 🚀
Kuse
@xi_z Hi Xi, really appreciate this thoughtful breakdown. You nailed it - we started from a simple canvas + LLM idea, but it was exactly those real user behaviors that pushed us into building a true unified workspace.
On your question about unexpected workflows: yes! A few that surprised us:
Educators dropping in PDFs + YouTube lectures, then using the canvas to auto-generate quizzes and lesson plans side by side.
Consultants mixing Excel financial models with research PDFs, then asking Kuse to generate strategy docs that pull from both.
Students building mind maps with convo cards + flashcards, then turning them into shareable study guides with Visualizer.
These weren’t in the original roadmap, but they showed us just how flexible the “canvas as context” can be.
We’re excited to keep evolving in that direction, and real-time collab is very much on the way!
PopPop AI Vocal Remover
Just give it a try! What is Kuse AI best suited for if I don't have a tech or coding background?
Kuse
@charlenechen_123 Thank you, Charlene! 😊 You don't need any tech or coding skills to get started with Kuse.
Many non-tech people like educators, marketers, and admin staff use Kuse to brainstorm ideas, visualize their work, and generate a variety of deliverables. I’d say that Kuse suits people without a tech background even better.
AgentX
If the product is built with tldraw, why is the tldraw watermark partially hidden in the corner? 🧐
Not sure if this is a mistake or if the license wasn’t paid.
Kuse
@douglas_lai Hey Yueming! Yeah it's a frontend glitch while we have been we’ve been sprinting hard lately. The tldraw logo shows properly in our 2.0 beta, and we’ll fix this soon. Appreciate you pointing it out!
Kuse
@rajat_jain33 Hey Rajat, you can adjust font size, color and add bullet points in "Text". We are on the way to make it happen for knowledge card as well.
Onlook
Kuse
@d_r_farrell Thanks! We are built upon @Tldraw. The sdk is pretty smooth. We rarely experienced with performance issue even with super large amount of objects on the canvas.