Jackal Labs is developing an event-driven analysis engine. We aggregate financial news, process sentiment using NLP, and backtest strategies against historical market data.
We are building pipelines to consume structured (OHLCV) and unstructured (RSS, API) data. Normalizing disparate feeds into a unified BigQuery schema.
Utilizing pre-trained Transformers (BERT/RoBERTa) to score financial news for sentiment polarity. Mapping text data to volatility events.
Simulating strategy performance against historical data to ensure statistical significance before any live deployment.
GCP Project initialization, IAM configuration, and basic data pipeline architecture.
Fine-tuning FinBERT on financial datasets using Vertex AI. Developing sentiment scoring API.
Live paper-trading environment deployment on Cloud Run. Real-time latency optimization.
We are building on industry-standard cloud technologies.