BethNa AI
Autonomous AI
The problem BethNa AI solves
Democratizing Institutional-Grade Options Trading with Swarm Intelligence
Crypto options trading allows for powerful hedging and yield generation, but it remains inaccessible to 99% of retail users. The barriers are high: complex terminology ("Greeks," IV, Strike Prices), the need for 24/7 market monitoring, and the intimidating interface of professional derivatives platforms. Retail traders often lose money not because of bad ideas, but due to poor execution and lack of risk management.
BethNa AI Trader solves this by abstracting the entire complexity of options trading into an Autonomous Swarm Agent System. Instead of forcing users to become derivatives experts, we provide them with a team of specialized AI agents:
- Agent Delta (Financial Advisor): Solves the "analysis paralysis" by onboarding users with a conversational interface, assessing their risk tolerance, and recommending strategies tailored to their profile (Conservative, Balanced, or Aggressive).
- Agent Alpha (Data Analyst): Solves the need for 24/7 monitoring by continuously analyzing market data, volatility, and technical indicators to find optimal entry points.
- Agent Beta (Executioner): Solves execution risk by automatically placing trades on Thetanuts Finance V4 via the SentientTrader smart contract, ensuring slippage protection and removing emotional bias.
- Agent Gamma (Social & Transparency): Solves the "black box" problem of AI by broadcasting every trade and rationale to Farcaster/Twitter in real-time, creating a fully transparent on-chain track record.
By combining these agents, BethNa converts a complex financial instrument into a "set-and-forget" experience where users maintain full custody while leveraging institutional-grade strategies on the Base L2 network.
Challenges we ran into
- Missing Testnet Infrastructure Thetanuts V4 wasn't available on Base Sepolia testnet yet, which blocked our testing.
- Solution: We built and deployed our own Mock Contracts that mimic the mainnet protocol perfectly. This allowed us to fully validate our agent's execution logic on-chain before launch.
- Synchronizing Multiple Agents Coordinating Agent Alpha (Python) and Agent Beta (TypeScript) in real-time caused state sync issues.
- Solution: We switched to an Event-Driven Architecture. Agents now listen directly to on-chain TradeExecuted events on Base. The blockchain became our reliable "single source of truth," eliminating sync errors.
- Performance vs. Visuals Our detailed "Glassmorphism" UI initially caused significant lag (High LCP).
- Solution: We implemented aggressive lazy loading for Web3 components and moved 3D animations to the GPU. This kept the UI buttery smooth (60fps) even while processing live market data.
Tracks Applied (1)
