Created on 26th June 2025
•
Customer support is often expensive, unreliable, or hard to set up for website owners, especially when trying to personalize answers based on their unique data. Existing chatbot solutions don’t offer:
Supportly solves this problem by providing a plug-and-play AI support system where users can upload their own data and pay only when customers ask questions using x402 microtransactions in USDC. This makes smart, personalized support affordable and easy to use. Under the hood, AI RAG (Retrieval-Augmented Generation) is used to ensure that responses are accurate and based on the user’s uploaded content.
Users can:
One of the biggest challenges was using x402 microtransactions throughout the system. Setting it up correctly and making sure every request was billed properly took a lot of trial and error. Thanks to the CDP team, they had well-written sample code on GitHub which really helped me understand and implement x402 correctly.
Deployment was also tricky. Setting up the backend (especially the payment mediator and AI model servers) and making them publicly accessible on the cloud took time, especially with firewall rules and service connections.
Another big hurdle was integrating AI with RAG (Retrieval-Augmented Generation). It was hard to make sure the AI models could understand and use the user-uploaded data correctly. I had to tune the flow, test different embedding and search setups, and make sure the right model was being used for each project.
In the end, I got through these challenges by reading docs, testing in small steps, and asking for help when needed.
Tracks Applied (5)
Amazon Web Services
Pinata
Technologies used
Cheering for a project means supporting a project you like with as little as 0.0025 ETH. Right now, you can Cheer using ETH on Arbitrum, Optimism and Base.