Mintara resolves the inefficiency of creating and distributing personalized tokens tied to social media interactions, specifically on X. Social media platforms generate vast user engagement data (likes, comments, followers), but lack effective mechanisms to leverage this data for targeted token distribution. Traditional community engagement methods often fail to identify and incentivize the right audience, leading to poor resource allocation and low impact.

Mintara enables the creation of tweet-linked coins using bonding curves on the Aptos blockchain. These coins incentivize users to engage with tweets, creating a viral loop of interaction while promoting campaigns. The system uses advanced graph traversal algorithms and retrieval-augmented generation (RAG) to analyze social connections and metadata. Each user is represented as a node in an on-chain graph database, with metadata stored as IPFS hashes. This metadata is embedded into an on-chain vector database, enabling semantic search to identify individuals most relevant to the campaign.

Once the initial audience is identified, Mintara uses graph traversal techniques (e.g., BFS) to expand reach by tagging individuals connected to these nodes. This ensures tweet-linked coins reach not only the primary audience but also secondary clusters likely to amplify engagement. By combining decentralized storage (IPFS), semantic search (vector similarity), and on-chain operations, Mintara creates a scalable framework for targeted token distribution.

Mintara incorporates GCP Trusted Execution Environments (TEEs) for secure backend operations, including the X bot that monitors tweets and triggers token creation. TEEs ensure sensitive operations are performed securely and isolated from threats. However, due to limited credits for maintaining TEE infrastructure (e.g., Letta) and X API access, some components are currently inactive. If awarded an Aptos grant, we can redeploy TEEs and secure sufficient API credits.