Really interesting project—always great to see more real-world applications being built on Polkadot. 
Since you mentioned this is your first Treasury proposal, I’d like to share a few suggestions that I’ve learned from following and participating in OpenGov over the years (We had been the Kusama DV and our proposal of AIWeb3 luckily had grant before, you can check our latest post here ):
• Start small. It’s often easier to build community confidence with a smaller proposal, deliver successfully, and then expand later.
• Define measurable milestones. Clear deliverables and timelines make it much easier for delegates to evaluate progress.
• Engage with the community early. Feedback on the forum before and during the referendum can significantly improve the proposal.
• Think beyond funding. Explain how your project creates long-term value for the Polkadot ecosystem, not just how the funds will be spent.
We’re currently taking the same approach with ComputeFlux. Rather than asking for a large Treasury proposal, we’re focusing on shipping an MVP, opening an early waitlist, gathering community feedback, and proving value first.
If your project ever needs AI tokens, we are happy to collaborate with you! Feel free to reach out!
Wishing you the best with the proposal!