TRIGGERSTOCK

A high-performance, real-time crypto analytics and prediction platform.
Built in Rust with full data pipelines, multimodal ML, and MLOps for intraday tick prediction.

---

🚀 Overview

• ⚡ Rust async pipelines (Tokio + Redis + Postgres) for low-latency ingestion
• 🌐 Finnhub WebSocket streaming ~160k+ rows/day at <10ms latency
• 🗄 Redis → Postgres bridge for persistence of OHLCV & trades
• 🧮 DAG-based feature engine (10+ TA indicators) generating >100k training samples
• 📰 Transformer-based sentiment analysis from Coindesk RSS, symbol-mapped
• 🤖 Multimodal signal fusion (price + volume + sentiment) via XGBoost + Transformers
• 📊 Grafana dashboards for real-time monitoring (latency, throughput, accuracy)
• ☁️ Northflank Blue-Green deployments with MLflow for experiment tracking & 99.9% uptime

---

🧠 What It Does

Component	Description
✅ ws_ingestor.rs	Connects to Finnhub WebSocket and streams live prices into Redis (<10ms)
✅ fetcher.rs	Periodically writes OHLCV from Redis into Postgres with TLS/NoTLS fallback
✅ news_ingestor.rs	Collects Coindesk RSS, maps to symbols, stores JSON headlines in Redis
✅ dag_engine.rs	Computes 10+ TA indicators (RSI, MACD, VWAP, etc.) for training datasets
✅ xgboost_trainer.py	Trains tick prediction classifier, logged via MLflow
✅ sentiment_model.py	Transformer-based sentiment scorer for financial news
🧪 deployment/	Blue-Green deployment scripts with MLflow model registry + Northflank CI

---

🏗 Architecture

flowchart TD
    A[Finnhub WebSocket] -->|Trades/OHLCV| B[websocket.rs]
    B --> C[Redis]
    C --> D[fetcher.rs]
    D --> E[Postgres: stock_price_history]
    C --> F[news_ingestor.rs]
    F --> C
    E --> G[dag_engine]
    C --> G
    G --> H[xgboost_trainer.py + sentiment_model.py]
    H --> I[MLflow Registry]
    I --> J[Northflank Blue-Green Deployment]
    J --> K[Grafana Monitoring]

Loading

---

⚙️ Tech Stack

• 🦀 Rust (Tokio, async, Redis, Postgres)
• 🐍 Python (XGBoost, Transformers, MLflow)
• 🗄 Redis + PostgreSQL (real-time + persistence)
• 📊 Grafana (real-time monitoring & dashboards)
• ☁️ Northflank (Blue-Green deployment, CI/CD, 90% uptime)

---

📈 Model Training

• XGBoost tick classifier with DAG-engine features
• Transformer sentiment fusion for news-driven volatility
• Metrics tracked via MLflow: accuracy, F1, false-signal reduction (~18%)
• Continuous retraining + deployment with zero downtime

---

🤝 Contributing

PRs are welcome!
If you're interested in real-time systems, crypto modeling, or ML infra, open an issue or contribute directly.

---

📜 License

MIT © 2025 Swastik Nandy