The problem BoomBox solves

Creating a new token traditionally requires technical expertise, visual design, branding, and on-chain deployment steps that are slow, repetitive, and inaccessible to most creators.

BoomBox eliminates this friction by allowing anyone to instantly create and deploy a token using only a sound file (voice, music, or video). It democratizes token creation by removing technical and creative barriers — no code, no design, just sound.

Behind the scenes, BoomBox utilizes AI to analyze audio content across multiple dimensions — including emotion, tone, sound type (e.g., dialog, music, fart), and semantic meaning. This data is then used to generate a unique token name, ticker, and token image that reflects the essence of the audio. Finally, the token is deployed trustlessly on Base via Clanker.

In short, BoomBox democratizes token creation and branding through AI — turning any sound into an instantly tradable, on-chain asset.

You can try out our XMTP agent on the XMTP Dev Portal via 0x34e3688f52f19912a4610f22c4a92d2aed5ae202

BoomBox is just getting started. We built our web client and Farcaster mini-app before the hackathon and developed our XMTP agent during the event.

Here’s what we’re building next:

1. Group Chat Support: We plan to support XMTP-based group chats, allowing communities to co-create and co-own sound tokens. This unlocks use cases like meme creation contests, collective remix drops, and community-voted naming flows — all done natively in chat.

2. In-Chat Utility Expansion: Beyond token creation, we aim to introduce richer in-chat utilities such as tipping with BoomBox tokens, querying token metadata, browsing recent creations, and triggering remix or burn actions — all directly within the XMTP conversation.

Challenges I ran into

• Ensuring accurate and relevant AI-generated metadata (name, ticker, image) from a wide variety of sound inputs.
• Building a reliable and seamless audio processing pipeline that handles YouTube links, X posts, and file uploads via XMTP or our web client.
• Integrating XMTP messaging in a way that allows users to tokenize audio directly from their DMs — including handling file attachments, parsing message context, and ensuring a smooth UX within a conversation.
• Managing the token deployment and metadata storage pipeline on Clanker & Base with low latency.
• Aligning user expectations with the output of generative AI when it comes to branding results (e.g., image style, name relevance).