Frequently Asked Questions

Predict & Profit 3.0: the full Python source for the Weather bot and the Inflation bot, the visual dashboard, the shared account-level risk engine, order and settlement reconciliation, the guided installer, the management CLI, diagnostics, migration and recovery tooling, and the customer documentation. The purchase is a personal, non-transferable licence.

Ubuntu 24.04 with Python 3.12 or 3.13 and PostgreSQL 16 or 17. Linux is the supported platform for running the application. Windows is fine as your browser or SSH client, and Windows users who want to run it locally should use WSL2. Native Windows lifecycle operation is not supported in 3.0.

You should be comfortable on a Linux command line, following setup instructions and reading an error message. The installer is guided and the documentation gives exact commands, but this is self-hosted source code you run yourself, not a desktop application with an installer wizard and a tray icon.

No. It is a one-time purchase of $129. There is no licence server, no metering, and the software does not phone home. You run it on your own infrastructure against your own Kalshi account.

No. It automates data collection, probability estimation, risk checks, order handling, accounting and settlement tracking. It cannot guarantee that a model is correct, that a market is mispriced, that an order fills, or that a trade makes money. Most scanned markets are expected to produce no trade at all.

No. Kalshi API access and the NOAA and Cleveland Fed data the bots rely on are free. FRED and BLS inputs are optional and also free.

Kalshi weather markets across 20 cities, plus CPI, Core CPI, PCE and related inflation markets. Polymarket is not supported, and neither is any other exchange.

The purchase includes six months of updates. Each update installs as a new immutable release beside your existing data rather than overwriting it. Setup support is available at support@predictandprofit.io, with responses within one business day.

3.0 ships migration tooling for supported earlier layouts. It inspects your existing installation, validates credentials and order ownership without modifying the source, converts the schema, and reconciles against Kalshi before anything is activated. Ambiguous ownership, missing history or precision loss blocks activation rather than guessing.

The software requires an eligible Kalshi account. Kalshi restricts access from some jurisdictions and may change its eligibility rules. Before purchasing, check the current Kalshi Member Agreement. If Kalshi does not permit access where you live, the trading features will not be usable.

Kalshi is a regulated US prediction market where you trade binary contracts on real-world events such as daily temperatures and macro releases. A trading bot automates scanning those contracts, estimating probabilities from data, and placing orders when configured conditions pass. Predict & Profit 3.0 includes one bot for weather and one for inflation.

It uses NOAA's National Blend of Models. NBM publishes calibrated, bias-corrected temperature guidance for the individual stations Kalshi uses for settlement, and the bot interpolates NOAA's published percentiles to estimate the probability of a contract threshold being crossed. NOAA GFS, NOAA AIGEFS, NOAA HRRR, ECMWF IFS and ECMWF AIFS remain in the model at a combined minority weight as a disagreement check.

Because most markets do not offer a position worth taking. The bot declines when forecast guidance is unavailable for the settlement station, when the spread or liquidity fails its controls, when the model and the market are close enough that the edge does not cover fees, or when a risk limit is already committed. A scan that places nothing is a normal outcome.

NBM coverage begins at forecast hour 24. Rather than fall back to a retired probability model, the bot declines to price same-day markets when suitable guidance is unavailable.

It builds a nowcast from the Cleveland Fed Inflation Nowcast and an internal weighted nowcast, with optional FRED and BLS inputs when those free API keys are configured, then compares that estimate against where the market is pricing the strike. It applies reference-month and settlement-aware checks, position-conflict detection, regime monitoring and the shared risk limits before acting.

Through a shared account-level risk engine. Both components trade one account under one set of limits, one account snapshot and one reconciliation path, so neither can spend the other's budget or leave the other with an inconsistent view of exposure. If the shared state cannot be resolved, both pause.

The bot takes fills, actual fill cost and fees from the authenticated order record the exchange returns, and subtracts them when settlement is recorded. If that evidence is unavailable, the row fails closed for reconciliation rather than assuming a fee.

In PostgreSQL, in dedicated databases with restricted runtime roles, under one permanent installation root alongside your configuration and logs. Upgrades add a new immutable release and reuse that state, so your history is not rewritten by an update.

Yes. Prediction-market liquidity is limited, and traders using similar logic can compete for the same prices and available contracts. No strategy is guaranteed to retain an advantage as markets and participation change.

Start in dry-run mode. Installation does not start live trading, and activation is a separate, explicit step. If you later enable live trading, use the smallest amount you can afford to lose while you verify authentication, market discovery, logging, settlement handling and risk limits in your own environment.

No. Predict & Profit is educational software for data science and research. Prediction markets are speculative, and you can lose the full amount committed to a position. You remain responsible for every trading decision and configuration change.