PRODUCT PREVIEWThis site showcases Orbit’s local development build. Screenshots and replays use example content. The app, accounts, uploads and live AI are not public yet.Availability →

Market research from your sources. Evidence in focus.

Feed Orbit the news, filings and notes you trust. It separates what the sources actually say into verified bullish and bearish evidence, tells you how strong that evidence really is, and shows you where it stops.

Not investment advice. Orbit cannot place trades, has no live market data and cannot predict prices. Sources can be wrong or out of date.

Markets mode showing verified bullish and bearish evidence for one instrument, with an evidence rating of mixed
Markets mode in the real app, using sample documents in demo mode.

How Markets mode works.

  1. 1

    Add your sources

    Attach articles, filings and notes. Orbit only reasons from what you give it.

  2. 2

    Evidence is verified

    Each claim needs an exact quotation, checked word-for-word. Unverified claims are not shown as evidence.

  3. 3

    Strength is computed

    Orbit, not the model, rates the evidence from insufficient to strong, and downgrades a stance the evidence does not support.

  4. 4

    Test it on paper

    Record simulated trades, backtest a simple rule on prices you supply, and size risk. No real money moves.

Two calculators that use Orbit’s own rules.

The same rules as the app, copied exactly and kept identical by an automated test. They run in your browser, and nothing you type is sent anywhere.

These interactive examples calculate locally in your browser. Use fictional numbers; nothing is sent or saved. They are arithmetic demonstrations, not investment recommendations.

Evidence strength

Enter what the verified quotations show. Orbit applies its rule and says why.

mixed

Evidence strength is not a probability and not a forecast. It rates how well the sources you gave support a view.

Position size from risk

Arithmetic on numbers you give. It is a calculator, not a recommendation.

    How the backtester avoids flattering the past.

    A backtest can look better than it is. These are the choices Orbit makes to keep it honest.

    Tuned on this partReported separately
    A diagram, not a result. The last 30% of the data is held back and reported on its own.
    • No look-ahead. A signal is computed at a close and traded at the next close.
    • Costs are charged. Fees and slippage come off every trade, and the cost drag is reported.
    • Out-of-sample is separate. The last 30% is shown on its own, so a rule tuned to the past cannot hide.
    • Long-only, five simple rules. Buy and hold, moving-average crossover, channel breakout, momentum and RSI mean reversion, on prices you supply.
    • Past is not future. Every backtest says so, because it is true.

    What Markets will never do.

    ✕

    Place orders

    There is no brokerage connection, no order endpoint and no order tool. A test asserts none exists.

    ✕

    Use live market data

    It works from documents and prices you provide. Nothing is fetched, and prices you import are as old as you made them.

    ✕

    Predict prices

    No forecasts and no probabilities. Evidence strength describes your sources, not the future.

    ✕

    Act as your adviser

    Orbit is not a licensed financial adviser and does not give personal investment advice.

    A closer look today. The working app comes next.

    Explore the local preview through its screenshots and examples. Public accounts, uploads and live AI are not available yet.