Founder building a visual-first collaborative AI workspace (learning in public)
Hey Product Hunt 👋
I’m Dushyant — founder & CEO of Genstellar.ai
I’ve been building products for a while, but over the last
months I found myself increasingly frustrated with how we interact with AI. The models are insanely powerful… yet the way we use them still feels oddly limiting. Linear chats, lost context, messy outputs, and very little room for real thinking or collaboration.
That frustration is what led me to start building Genstellar — a visual, spatial AI workspace designed to match how humans actually think, explore ideas, and work together. Still very much in the building phase, learning constantly, and refining things based on real feedback.
I joined Product Hunt because I genuinely enjoy the builder conversations here — especially around vibe coding, AI workflows, and the realities of turning prototypes into real products. I’m here to learn, share lessons (including mistakes), and contribute wherever I can.
Would love to connect with:
• founders building in public
• people experimenting with AI-first workflows
• anyone thinking deeply about the future of AI
What are you currently building — or thinking about building?
Looking forward to learning from this community 🙌
Replies
Thanks a lot for the thoughtful replies here 🙏
One thing I’m intentionally doing this time is building in public and pressure-testing ideas early — especially around how people actually use AI day-to-day, not just what looks impressive in demos.
While building Genstellar, I realized the real problem with AI isn’t intelligence — it’s how fragmented our thinking becomes when context, memory, and collaboration break down.
Some days it’s linear chats limiting exploration.
Other days it’s losing long-term context.
Sometimes it’s not being able to see ideas clearly or work on them with others.
Genstellar is my attempt to rethink AI as a visual, memory-aware workspace — where conversations can branch, context stays controlled (not bloated), ideas are searchable, and thinking feels more natural instead of forced.
This is all pre-launch, and I’m here to learn.
If you’ve struggled with AI memory, context overload, visualizing ideas, or collaborating around AI outputs — I’d genuinely love to hear how you deal with it today (or what’s still broken).
No pitching, just building alongside people who actually use these tools.
@dushyant_khinchi hey dushyant, just a quick question: what's your reason behind emphasizing visual over textual or audio? Let's say in case of a phone call session to brainstorm with AI, are visuals a priority?
@tinnix_he Great question — and yeah, I should clarify this because “visual” can sound limiting.
Genstellar isn’t visual-only and it’s not replacing text or audio.
Think of it more as everything becoming a first-class object on a canvas.
AI responses, notes, charts, dashboards, audio clips, images, even videos - they all exist as cards on a free-floating board. Text and audio are still there; they’re just represented visually so you can work with them.
The big advantage over a normal chat isn’t just a user intuitive workspace, it’s context control.
In a normal chat (or even a phone-style brainstorming session), thinking is forced into a linear back-and-forth.
If you see any AI tool out there today, whether they are text based, or audio based, or even for image and video processing - They all follow a sequential path. What are the limitations of that?
Context keeps growing, drifting, and eventually the model starts guessing what matters. That’s where hallucinations and “why did it say that?” moments creep in.
On a canvas:
You decide what stays in context and what doesn’t
You can branch ideas instead of overwriting them
You can group, isolate, or revisit thoughts without restarting
The AI reasons over selected context, not the entire conversation history
So even with minimal prompts, outputs tend to be sharper because the context is explicit, not implied.
Visuals here aren’t the goal — they’re the interface for thinking non-linearly, the way humans actually brainstorm. Audio and text are still powerful inputs, but the canvas is what keeps the thinking grounded instead of collapsing into one long chat thread.
And the best part is, you can still choose to have a traditional sequential back-and-forth brainstorming session with AI, but you are not bound to do so.
Hope I was able to make sense here. Do let me know if you have any thoughts or questions😊
@dushyant_khinchi would be fun to have a drawing board doodling with AI and brainstorming together someday!! literal canvas!
@tinnix_he Love that idea - and honestly, you’re not far off from where this is heading 😊
We’re actually about to roll out Skills inside Genstellar, which are like focused, task-specific AI modes that plug into the canvas. A doodling + co-brainstorming skill fits perfectly into that vision - imagine sketching rough ideas, flow diagrams, or concepts while the AI understands, reacts, and builds alongside you on the same board.
That’s exactly the kind of human-AI collaboration I’m excited about: not just typing prompts, but thinking together in a shared visual space.
Definitely adding this to the list for the upcoming skill rollout. Would love to hear what you’d want to doodle or brainstorm first on a canvas like that 👀