AIXBT Sigma by Buffett

The problem it solves

Trading profitably onchain is tough. Staying ahead requires: ✅ Being terminally online—tracking hundreds of Telegram chats, Discord channels, and Twitter KOLs. ✅ Staring at charts all day to time the market.

AIXBT flipped this meta by automating alpha discovery—curating high-signal trading insights that traders rely on. But there’s a new problem: 🚀 AIXBT tweets about ~900 tokens daily. 📉 It’s humanly impossible to trade at that scale manually.

That’s where AIXBT Sigma agent comes in. It analyzes AIXBT’s tweets, extracts trading insights, and executes buy-and-sell decisions onchain—all without human intervention. In our simulation over a three week period, Sigma managed a win rate of 40% with a profit of 80% (ETH price has fallen down by 2.45% during the same period)

Challenges we ran into

Twitter Scraping: Extracting real-time data is difficult due to strict API restrictions and rate limits, making it challenging to track new token mentions efficiently. Token Validation: Identifying valuable tokens among a flood of noise requires precise parsing, sentiment analysis, and rigorous checks to filter out spam tokens. Many mentioned tokens may exist on other chains, so cross-checking contract addresses to ensure they are native to Base is essential. Sell Logic Optimization: Finding the right exit strategy is complex and requires extensive backtesting and simulations to refine strategies for maximizing profits while minimizing risks.

Note to judges & Organisers: AIXBT Sigma is built on top of Buffett which is in invite only access currently. Please DM us on twitter for access to our product & github