Skip to content
View TPTBusiness's full-sized avatar
18 followers · 102 following

Block or report TPTBusiness

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
TPTBusiness/README.md

Trading Prediction Technology · TPTBusiness

Building AI-powered trading systems — from autonomous factor discovery to local LLM inference.
Supporter of open-source software and the communities that build it.

🚀 Main Project — Predix

Predix is an autonomous AI agent for quantitative EUR/USD forex trading. It automates the full research and development cycle — from factor discovery to backtesting — using a multi-agent LLM framework on 1-minute data.

What makes it different:

  • 🧠 Autonomous factor evolution — the agent proposes, codes, and validates its own alpha signals
  • 🛡️ Built-in risk management — drawdown protection, cooldown periods, stoploss clustering detection
  • 🔄 Walk-forward validation — avoids overfitting across 2020–2026 EUR/USD data
  • 🖥️ Real-time dashboard — Streamlit UI for monitoring factor performance and model evolution
  • 🔒 60 integration tests — every commit is checked before it lands
 
Market Data (CCXT / yfinance)
         ↓
  Feature Engineering
   (TA-Lib, pandas)
         ↓
  LLM Agent Loop
  (factor proposal
   + code generation)
         ↓
   Backtesting
   (Qlib, 1-min)
         ↓
  Risk Management
  (drawdown / cooldown)
         ↓
  Portfolio Output

GitHub License Last Commit Stars

🧠 What I build

Autonomous Trading Agents — Multi-agent LLM frameworks that discover, evolve, and validate trading strategies end-to-end
Local LLM Integration — Running AI systems fully offline with llama.cpp (no cloud dependency)
Open-Source Tools — Pine Script strategies and Python frameworks for the trading community
Full Trading Pipelines — From raw kline data to live execution, built and maintained independently

🗺️ Predix Roadmap

Status Phase Feature
✅ Done P0 Data Loader — OHLCV loading, HDF5 caching, thread-safe feature matrix builder
✅ Done P1 Strategy Worker — LLM call wrapper, backtest engine, FTMO compliance gate
✅ Done P2 Strategy Orchestrator — multi-process pool, LLM semaphore, result deduplication
✅ Done P3 Optuna Optimizer — TPE sampler, FTMO penalty logic, 20–50 trials per strategy
✅ Done P4 CLI Commands — generate_strategies, Rich console output
✅ Done P5 ML Training Pipeline — LightGBM, time-series split, feature importance analysis
✅ Done P6 fin_quant Feedback Loop — ML feature importance → LLM prompt feedback
✅ Done P7 Portfolio Optimizer — mean-variance, risk parity, Black-Litterman with LLM views
✅ Done P8 Integration Tests — end-to-end pipeline, parallelisation, FTMO compliance
✅ Done P9 Documentation — architecture diagrams, setup guide, data flow
✅ Done P10a Kronos-mini — OHLCV foundation model inference on EUR/USD (4.1M params, AAAI 2026, MIT)
📋 Planned P10b Kronos as factor generator — kronos_predicted_return_96, volatility, momentum, uncertainty
📋 Planned P10c Kronos + LLM Ensemble — Optuna-optimized weighting of DL + LLM alpha signals
📋 Planned P10d Kronos fine-tuning on EUR/USD 1-min custom tokenizer (optional)
📋 Planned P11 Live execution — offline/online split, Telegram signal alerts

🛠️ Stack

Core & AI

Python PyTorch LightGBM XGBoost scikit-learn LangChain

Model Architectures

Transformer LSTM PPO Stable-Baselines3

Data & Finance

pandas NumPy TA-Lib CCXT Qlib yfinance Binance API

Local LLM & Inference

llama.cpp Ollama CUDA OpenRouter

UI & Infra

Streamlit Docker Linux Pine Script

🌍 Open-Source Contributions

Project Contribution
TradingAgents ⭐ 34k Added llama.cpp local LLM support — run multi-agent stock analysis fully offline via .env config
OpenStock ⭐ 9.9k Updated deps, fixed Inngest v4 API, force-dynamic for auth routes — resolved 28 vulnerabilities, migrated Inngest v3→v4

📊 Activity

GitHub followers GitHub stars Predix commits Predix issues closed

📬 Contact

Premium models & collaborations → [email protected]
Mastodon → @[email protected]

⚠️ All content is for educational purposes only. Past performance does not guarantee future results.

Pinned Loading

  1. Predix Predix Public

    Autonomous AI agent for quantitative EUR/USD forex trading — automates factor discovery, model evolution, and backtesting on 1-minute data using a multi-agent LLM framework.

    Python 2

  2. VAMOStrategy VAMOStrategy Public

    Open-source Pine Script v6 strategy for TradingView, optimized for Bitcoin scalping. Uses a Volatility-Adjusted Momentum Oscillator (VAMO) for crossover signals, with ATR-based stops and profits, a…

    4