Questions serious traders ask before they trust a new research workflow.

PredictDesk helps you turn a trading question into a research workflow: define the symbol, structure, timeframe, and goals; scan the setup space; inspect qualified candidates and near misses; then carry the full setup into testing, Builder, monitoring, or webhook routing. It is research tooling, not advice or execution.

Screeners usually filter symbols. Strategy Lab starts with the setup you want to test and searches combinations around that setup, including structure, entry logic, exit logic, trade management, risk, and goal criteria. The output is a set of strategy candidates for deeper review, not signals to chase.

A qualified candidate cleared the goals you set for that scan, such as expectancy per 100 trades, trades per year, and P95 trade risk, based on available historical data and assumptions. It still needs trader review and is not a recommendation.

Near misses are candidates that came close but missed at least one goal. They help you see what held an idea back, such as frequency, risk, expectancy, filters, or robustness, so the scan teaches you more than a simple pass/fail result.

You can inspect the evidence, review the full configuration, save the candidate, carry it into Builder, run hypothetical backtests or walk-forwards, and decide whether it deserves monitoring. The setup context stays attached instead of forcing you to rebuild the idea from scratch.

Yes. PredictDesk supports stock and options research paths in beta, including underlying workflows and options structures such as debit spreads, credit spreads, and iron condors where supported by the current beta surface.

Yes. Backtests and scenario outputs are included as hypothetical research context. They help pressure-test assumptions and compare candidates, but they do not predict returns or guarantee future performance.

Walk-forward review helps test whether a candidate remains coherent across sequential historical windows instead of only looking strong in one fitted sample.

P95 trade risk estimates the 95th-percentile loss per trade in R terms for the tested setup. It helps compare candidates against a trader-defined risk bar, but it remains model, data, and assumption-bound.

Diagnostics show which goals cleared, which nearly failed, what filters or conditions mattered, and whether the candidate looks robust enough for more work. They are meant to support review, not replace judgment.

Coverage depends on symbol, instrument, interval, and available data. Options and intraday paths may have narrower coverage than simple daily equity paths, and the docs disclose the relevant horizons and assumptions.

Yes. Results can change when data coverage, assumptions, filters, date windows, contract construction, or market conditions change. Treat every output as research evidence, not a permanent answer.

Builder turns a saved candidate into a configurable strategy workspace. You can adjust logic, create versions, add guardrails, select endpoints, and prepare monitoring context without losing the original research setup.

Yes. PredictDesk can monitor configured strategy events and show delivery state in Signals Feed. Monitoring context helps you see what happened; it does not decide, place, or manage trades.

You add endpoint destination URLs, then attach an endpoint in Builder when a strategy is ready for delivery monitoring. Endpoint events are outbound notifications you configure. They are not broker execution.

Research includes up to 6 outbound webhook endpoints. Pro includes up to 50 outbound webhook endpoints. Desk is contact-led for team and firm workflows.

PredictDesk is broker-agnostic. Broker order entry, account management, fills, positions, and custody remain outside PredictDesk.

Signals Feed shows delivery events such as accepted, failed, cancelled, or suppressed states, with endpoint and strategy context so you can monitor routing health.

Research is for individual traders who want Strategy Lab, diagnostics, backtests, saved configurations, methodology docs, and a limited number of outbound webhook endpoints.

Pro is for heavier individual users who want more monitoring headroom, deeper workflow capacity, richer diagnostics, and more outbound endpoint capacity.

Desk is for teams, firms, and desks that need shared workflows, permissions, governance, onboarding support, and a guided commercial access path.

No. PredictDesk is self-directed research tooling. It does not recommend trades, provide personalized investment advice, manage accounts, custody assets, or guarantee outcomes.

PredictDesk is available through invite-only beta. Access is granted by cohort and workspace readiness. Joining the waitlist or requesting access records interest; it does not guarantee access.

Billing begins only after you choose a plan or approved access path and accept the applicable terms. Plan and cancellation details are shown before checkout or agreement.