PredictionToolkit logoPredictionToolkit
Language: English
Search tools
← AI agents

Research tool

Prediction Pilot: research and scan Kalshi strategies in chat

Prediction Pilot combines market data, NWS forecasts and web sources in cited research. It turns plain-English strategy descriptions into saved scan rules, tests them on historical Kalshi data and monitors positions through read-only access.

3.5/ 5Editorial ratingEditorial assessment
Open Prediction Pilotpredictionpilot.io
Available on
Web
Availability: public

Screenshots and interface

Product preview1 / 2

The backtesting view introduces historical strategy evaluation and diagnostics.

What is Prediction Pilot?

Kalshi is the primary venue. Relevant Polymarket contracts serve as cross-checks for equivalent-contract prices, rather than providing full feature parity. The assistant does not place trades automatically; users review a proposed opportunity and click through to Kalshi themselves.

Backtests use a described 90-day settled-market dataset with training and validation splits. Trade count, PnL, drawdown and category results help evaluate a rule, but that history may be a small sample for infrequent strategies.

Who it is for

  • Kalshi users seeking chat-based research with sources.
  • Researchers testing simple rules before scheduling scans.
  • Users monitoring exposure and equivalent-contract price gaps.

Features and coverage

  • Cited opportunity research

    Combine venue data, NWS forecasts and web sources while inspecting volume, spread, closing time and price bands.

  • Rules and historical diagnostics

    Convert a natural-language setup into saved conditions and backtest on 90 days of settled Kalshi data with train/validation splits, trade counts, PnL, drawdown and category results.

  • Scheduled scans and notifications

    Run saved rules every 5 minutes, hourly or daily and receive push, email or in-app alerts.

  • Read-only position monitoring

    Connect Kalshi read-only data to review exposure and position flags without delegating trade execution.

  • Polymarket cross-checks

    Compare relevant contracts during research, checking outcome equivalence, book depth and fees before interpreting a price gap.

Workflows with Prediction Pilot

Test a described rule

Inspect the conditions produced from the prompt, then assess sample size and drawdown alongside returns.

Monitor a strategy

Save the rule, choose its schedule and notification channel, and review each match yourself.

Investigate a price gap

Check contract equivalence, current depth and fees before opening either venue; the assistant does not execute the trade.

Pricing and total cost

A free, no-card signup allowance is followed by a credit-based monthly plan. Message estimates are not a fixed allowance: usage depends on model and question size.

Signup allowance

Free

  • 250,000 credits
  • No card required

Copilot

$14/month

  • 2,500,000 credits
  • Research, backtests and scheduled scans
  • Position monitoring

Fees and usage costs

  • Overage costs $10 per million credits.
  • The FAQ lists default soft/hard overage caps of $10/$25, with configurable warning and pause caps.

Getting started

Before you start: Kalshi is primary; Polymarket support is limited to cross-checking.

  1. Start with the free credits and ask a research question, then inspect its sources.
  2. Connect Kalshi read-only access if portfolio analysis is needed.
  3. Test the strategy before scheduling scans.
  4. Configure notifications and usage caps.
  5. Review proposed trades and execute them yourself in the venue.

The assistant does not place trades automatically.

Useful links

    What to check

    • Kalshi is primary; Polymarket support is limited to cross-checking.
    • The 90-day backtest window may be too small for infrequent strategies.
    • Marketing message estimates vary with model and question size.
    • Scheduled scans and position monitoring do not authorize automatic orders.

    A closer look

    Ratings & reviews

    Editorial rating
    3.5/ 5
    0 user reviews
    PredictionToolkit editorial

    Our take on Prediction Pilot

    Cited research and testable scan rules create a clear assistant workflow while leaving execution with the user. Kalshi’s primary role, narrower Polymarket comparison and credit consumption are central to judging recurring use.

    Based on documented features, setup and access terms. How we rate

    User reviews 0

    The main score switches to user ratings after 5 published reviews.

    Be the first to share your experience

    A specific example helps the next person choose the right tool.

    Related tools

    Browse category
    Octagon product overview.
    Octagon
    Research tool3.5Editorial

    Source-linked event reports, model-price comparisons and developer access through API, MCP and CLI.

    • Access Web · API · CLI · MCP

    Compare another AI research workspace.

    CloddsBot repository overview with the provider’s embedded WebChat preview
    CloddsBot
    Research tool3.0Editorial

    Run market research and configured strategy tools through a local chat gateway and supported messaging channels.

    • Access Web · Telegram · Discord · MCP

    Also in AI Agents and Assistants.

    Miramarket product overview
    Miramarket
    Trading terminal3.0Editorial

    Build and inspect conditional trading rules on a canvas, then simulate before choosing live execution.

    • Trading mode Paper · Live automation
    • Access Web

    Compare visual strategy automation.

    Frequently asked questions about Prediction Pilot

    Does it execute orders?

    No. The product FAQ says users must click through to Kalshi themselves.

    Is a card needed to try it?

    No. The starting credit allowance is advertised without a card requirement.

    Can a chat request run a backtest?

    Yes. The described backtesting workflow uses settled Kalshi market data and includes diagnostic results.

    Can I limit extra usage charges?

    The product describes configurable warning and pause caps. Review them alongside the included credits and overage terms.