SparklingAI

FAQ

Frequently asked questions

Common questions about SparklingAI, AI trading research, walk-forward testing, XAUUSD backtesting, and the limits of public research notes.

1. SparklingAI

  1. 1. Is SparklingAI only for gold trading?

    No. Gold is the first public research track. The broader goal is AI market intelligence across more assets over time.

  2. 2. Is SparklingAI already a SaaS or API product?

    No. The current site is the research layer. A SaaS dashboard, API, or subscriber product may come later after more validation.

  3. 3. Why explain the stack without publishing the full model?

    Public research can explain the architecture, validation process, and risk philosophy without exposing proprietary alpha logic. That balance keeps the content useful without giving away the future product.

2. Walk-Forward Testing

  1. 4. What is walk-forward testing in trading?

    Walk-forward testing is a validation method that trains or tunes a strategy on one period and tests it on a later unseen period.

  2. 5. Why is walk-forward testing useful for AI trading?

    It helps expose overfitting by testing whether a model still behaves well outside the data used during development.

  3. 6. Does a profitable fold prove the strategy works?

    No. A profitable fold is evidence, but the full fold sequence matters more than one isolated result.

3. XAUUSD Backtesting

  1. 7. What is XAUUSD backtesting?

    XAUUSD backtesting is the process of testing a gold trading strategy on historical gold price data before risking real money.

  2. 8. Why use walk-forward testing for XAUUSD?

    Walk-forward testing helps separate a strategy that only fits the past from a system that can survive multiple out-of-sample market windows.

  3. 9. Does the +14.15% result prove the system is ready for live trading?

    No. The rerun is encouraging, but it is still a research result. Live trading adds more uncertainty, including execution quality, spread changes, slippage, latency, broker behavior, and future market regimes.

  4. 10. Why not publish the full alpha logic?

    The public article is meant to show the validation process. The alpha construction, thresholds, feature recipes, and execution rules remain private because they may become part of a future SparklingAI product.

4. Risk And Disclaimer

  1. 11. Is this financial advice?

    No. SparklingAI content is for research and education only. It is not financial advice, investment advice, trading advice, or a recommendation.

  2. 12. Do backtested results guarantee future performance?

    No. Backtested or simulated results do not guarantee future performance. Live trading includes market, execution, broker, spread, slippage, and liquidity risk.

Read the full disclaimer before using any research note as context.

Open disclaimer