Video unavailable. Open the source below.
NewScarlett Trading Analysis
An experimental market-analysis demo combines price forecasts with Jev decisions.
Added to Jevfast
How it uses Jev
A separate autoregressive model forecasts price movements. Jev evaluates a feature space containing time-series data, those forecasts, technical indicators, and news to identify trading opportunities as conditions change.
App workflow
- Forecast prices
An autoregressive model trained on 12 billion data points supplies price-movement forecasts.
- Assemble market features
The decision feature space includes time-series data, forecasts, technical indicators, and news feeds.
- Evaluate opportunities
Jev makes decisions over those features as market conditions change.
- Surface candidates
The demo presents opportunities for Hyperliquid; the supplied post does not detail execution.
Why it is interesting
The design separates forecasting from bounded opportunity decisions and combines model output with market indicators and news. The post describes an experimental workflow, not verified trading performance.