BuffiCast

About

The problem it solves ETHDenver is an amazing event, but the huge volume of talks, workshops, and parties can be overwhelming. Attendees struggle to navigate the schedule and often miss out on valuable experiences. BuffiCast solves this by providing curated, AI-powered podcast summaries, helping you discover the best of ETHDenver without losing hours scrolling and experiencing FOMO. BuffiCast acts as your personalized ETHDenver guide, eliminating the need for endless, manual research. We pull information directly from the ETHDenver Discord and then use our ElizaOs framework to curate it using on-chain randomness (powered by Flow). This ensures authentic and unbiased recommendations, helping you: Cut Through the Noise: Focus on what matters most by filtering out what's most relevant. Discover Hidden Gems: Uncover unique and last minute alphas. Stay Informed FAST: Get concise podcast summaries of key discussions without wading through endless replies. Authenticity: Our use of Flow's on-chain randomness ensures recommendations are based on unbiased data, not just popularity. Potentially Earn: Our programmatic IP approach opens up opportunities for revenue generation. In short, BuffiCast helps you navigate the ETHDenver chaos, stay informed with authentic and engaging content, and even potentially earn while doing it. We turn the noise into alpha, empowering you to make the most of your ETHDenver experience. User Interaction and Data Flow Discover BuffiCast NFTs: Find BuffiCast podcast episodes represented as NFTs through our community channels later accessed by a dedicated Bot that you can add to your Discord Server. Access Podcasts via NFTs: Listen to BuffiCast episodes directly through the associated NFT. Future Tokenized Interaction: In the future, users may interact more deeply with the tokenized podcasts (IPAssets), potentially unlocking features like exclusive content or governance participation. Data Flow: Daily Discord Scrape: Our system automatically scrapes the ETHDenver Discord channel daily, extracting key information and discussions. AI Curation & Randomization: This data is processed by our ElizaOs framework, which uses AI to curate content based on predefined criteria. Flow's on-chain randomness ensures diverse podcast topics. IPAsset Creation: The curated content is packaged into a podcast episode and minted as an IPAsset through Story Protocol. Podcast Distribution: The IPAsset is then distributed to our community. The project architecture and development process Overview: BuffiCast is a decentralized, AI-powered podcast creation and distribution platform designed to help users navigate the complexities of ETHDenver. It leverages the ElizaOs framework for AI agent capabilities, the Flow blockchain for on-chain randomness, and Story Protocol for programmatic IP management. Core Functionality: Data Acquisition: Daily scraping of the ETHDenver Discord server using the Discord API. Content Curation: The ElizaOs agent is initialized with predefined curation criteria (e.g., tone, narrative, unexpected twist). The agent processes scraped Discord data, identifying relevant discussions, events, and announcements. Flow's on-chain randomness selects specific traits and topics for the podcast, ensuring variety and unbiased curation. Podcast Generation: The AI agent creates a summary of the extracted information, creating a podcast script. The script is converted into audio using the Eleven Labs Text-to-Speech API. IPAsset Tokenization: Each podcast episode is minted as a unique IPAsset using Story Protocol's smart contracts. Metadata associated with the IPAsset includes the episode title, description, date, and other relevant details. Distribution: The IPAsset is registered on the Story Protocol marketplace. Product Integrations ETHDenver Discord API: Extracts real-time information, announcements, and discussions from the ETHDenver Discord server daily. a16z's ElizaOs Framework: Provides the foundation for building on-chain AI agents with capabilities for data processing, curation, and decision-making. Flow Blockchain: Enables on-chain randomness for unbiased content curation and smart contract execution for IPAsset management. Story Protocol: Facilitates the creation, management, and monetization of IPAssets representing individual podcast episodes. Eleven Labs Text-to-Speech API: Converts podcast scripts into natural-sounding audio. IPFS (via Piñata): Provides decentralized storage for IPAsset media. Key differentiators and uniqueness of the project BuffiCast distinguishes itself from other ETHDenver information sources through its unique combination of features: AI-Powered, On-Chain Curation: We leverage the ElizaOs framework to create an AI agent that actively curates content based on predefined criteria, providing a more personalized and relevant experience. Authenticity Through Randomness: By integrating Flow's on-chain randomness, we ensure that the podcast content remains diverse, unpredictable, and free from bias. Programmatic IP with Story Protocol: Tokenizing podcast episodes as IPAssets through Story Protocol allows for new revenue models and community engagement mechanisms. Real-Time Information: Our daily Discord scraping ensures the podcast content is always up-to-date. Combining Story Protocol and ElizaOS: Showcases a novel approach to building decentralized content curation tools. While some projects may offer ETHDenver event listings or news summaries, BuffiCast is unique in its integration of AI-powered curation, on-chain randomness, and programmatic IP, creating a more authentic, engaging, and sustainable ecosystem for both creators and listeners. Trade-offs and shortcuts while building During the BUIDLathon, we faced several trade-offs to meet the deadline: Limited AI Training: We used pre-trained AI models within the ElizaOs framework rather than training a custom model specifically for ETHDenver content. Future Plan: Train a custom AI model to better understand the ETHDenver community. Basic UI/UX: The user interface is currently basic. Future Plan: Develop a more user-friendly interface with features for searching, filtering, and managing IPAssets. Prioritized Core Functionality: We focused on core functionality (getting TL;DR from ETHDenver) over advanced features. Future Plan: Implement user accounts, personalized recommendations, and community interest features. We acknowledge these trade-offs and are committed to addressing them in future iterations to improve the overall experience of BuffiCast. Additional Features BuffiCast was conceived and built entirely during the ETHDenver BUIDLathon. Specifically, we: Integrated the ElizaOs framework for on-chain AI agent capabilities. Implemented Flow's on-chain randomness for unbiased content curation. Tokenized podcast episodes as IPAssets through Story Protocol. Developed the Discord scraping mechanism. Created the basic UI for accessing the podcasts. Connected Story Protocol and ElizaOS. Add the ability to build this on a crowded Discord