Today we rolled out a major update to FactorBench: a chat-first workflow that turns plain-English prompts into strategy screens and backtests. If you have ever wished for a ChatGPT-style interface for quant research, this is it.
What changed
- AI chat as the primary interface: Ask for a strategy, a screen, or a backtest in one message.
- Strategies and backtest configurations are separate: Build once, then reuse backtest settings across strategies.
- Defaults are smart: If you do not specify backtest settings, we run the default configuration for your plan.
- Results are saved: Screen and backtest results are stored in their own tabs for quick review.
Why it matters
This rework makes research faster and more repeatable. You can iterate on strategy logic without touching backtest settings, and you can rerun backtests with a single click from the chat.
Try it now
Open a chat, describe your idea, and run a screen or backtest in minutes. Start here.