β GPT model trained on Jesseβs voice and context (Farcaster + 160+ podcasts/videos)
β Adding knowledge base with no-code UI
β Using Puppeteer to scrape websites and make summaries
β GitHub evaluation for grants selection
β Connection to Warpcast, X, and Telegram profiles
π΅ Phase 1: Base Batches LATAM (Apr 25 β May 16)
β Finalize Gemini 2.5 fine-tuned model (Farcaster + podcasts/videos)
β RAG implementation using Pinecone
β Set up vector database (Pinecone) with namespaces (jessexbt, builders, protocols)
β Puppeteer site scraping with automatic/manual refresh
β Integrate public data ingestion (Farcaster, Twitter, Telegram) with retraining
β Auto-enrich GitHub, demo (Gemini 2.5), X threads
β Implement natural language prompt flow for builder/project evaluation
π’ Phase 2: Core Infrastructure and Alignment (Day 1β20)
β JesseXBT becomes a 24/7 Base-native agent
β Discovery engine for talent, ideas, and protocols
β Transparent builder scoring and protocol intelligence
β Trusted filter between Jesse and the noise
β Programmable assistant with memory, personality, and Base-native cognition