Platform Features
STOIC aims to democratize algorithmic trading and crypto analytics by combining the power of LLMs with community-driven insights on a scalable blockchain infrastructure. By integrating Stoics proven technology, providing robust backtesting and paper trading options, and building a vibrant marketplace around proven strategies, STOIC aspires to become a cornerstone in the rapidly expanding digital asset ecosystem.
Platform Features
Stoic AI has four core components:
No-Code Strategy Builder
Users describe a strategy in plain language (e.g., “Buy RSI below 30 and sell when RSI above 70”). Integrate a Large Language Model (LLM) that interprets user prompts and auto-generates code snippets for Freqtrade-based strategies.
The system interprets the description, generates the underlying trading logic, and visualizes it.
Users can adjust parameters (timeframe, indicators, risk settings) through sliders or dropdowns.
Backtesting Engine
Runs strategies on historical data from major exchanges (Binance, OKX, Bybit, Hyperliquid, Gate.io).
Data is sourced directly from each exchange’s official APIs or archives.
Results include profit/loss curves, drawdowns, and key metrics (Sharpe ratio, win rate).
Users can compare multiple versions of a strategy side by side.
Live Deployment Module
Connects to user exchange accounts via API keys.
Deploys strategies in a containerized environment on Kubernetes.
Monitors positions, orders, and risk in real time.
Provides a dashboard showing live performance and logs.
Marketplace
Strategy creators publish their models with descriptions and performance stats.
Buyers or subscribers can browse by category, performance, or fee structure.
Stoic AI handles subscription payments and access control.
Creators set their own prices; Stoic AI takes a 2–5% commission per sale or subscription.
4. Core Components in Detail
4.1 No-Code Strategy Builder
Natural Language Processing
Users enter plain‐language descriptions of buy/sell rules.
The NLP engine maps keywords (“RSI,” “MA,” “crossover”) to technical indicators.
It proposes a draft strategy script in Python/Freqtrade format.
Visual Interface
Drag‐and‐drop blocks represent conditions, indicators, and actions.
Users can link blocks to define logical flows (e.g., “If RSI < 30 AND price > 20‐day MA, then buy”).
Parameter panels let users set values for lookback periods, risk %, position size.
Templates & Examples
Built‐in templates for common strategies (momentum, mean reversion, breakout).
Example scripts show best practices for risk management.
Users can clone and tweak examples.
4.2 Backtesting Engine
Data Sources
Official OHLCV (open, high, low, close, volume) feeds from each exchange.
Minute‐level and hour‐level data stored for up to two years.
Automatic updates every hour to keep the dataset fresh.
Simulation Framework
Uses Freqtrade’s backtesting module.
Supports backtesting on multiple timeframes (1 m, 5 m, 15 m, 1 h, 4 h, 1 d).
Accounts for fees, slippage, and order types (market, limit).
Result Metrics
Net profit and loss (PnL).
Maximum drawdown.
Sharpe ratio.
Win rate and profit factor.
Charts: equity curve, drawdown curve, distribution of returns.
Reporting
PDF or CSV export of detailed logs.
Side‐by‐side comparison of up to five strategies.
Downloadable backtest configurations.
4.3 Live Deployment Module
Exchange Integration
Support for Binance, OKX, Bybit, Hyperliquid, Gate.io.
Users generate API keys with read/write permissions and input them in Stoic AI.
Secure key storage: encrypted at rest and in transit.
Containerized Strategy Runners
Each user deployment spins up a separate pod running Freqtrade.
Pods auto‐restart on failure.
Health checks monitor order placement and connectivity.
Monitoring & Alerts
Real‐time dashboard shows open orders, filled orders, PnL.
Logs streamed to Cloud Logging.
Alerts for connectivity issues, API rate limits, or failed orders.
4.4 Marketplace
Listing Strategies
Creators submit a name, description, and performance summary.
They upload backtest reports and any supporting documentation (PDF, screenshots).
They select a pricing model: flat fee or subscription.
Stoic AI verifies the backtest report for data validity but does not guarantee performance.
Browsing & Discovery
Users filter by asset class (e.g., BTC, ETH, altcoins), timeframe, risk profile, or price.
Sorting by highest return, lowest drawdown, popularity, or newest.
Ratings and reviews from subscribers.
Purchase & Subscription
Buyers pay with SOL or Stoic tokens.
Stoic AI holds payment in escrow for a trial period (e.g., 7 days).
If the buyer keeps the strategy live after the trial, funds release to the creator minus a 2–5% fee.
Creator Dashboard
View subscription counts, revenue, and churn rate.
Options to update or delist strategies.
Analytics on user engagement and feedback.
Last updated