Same process every project, regardless of venue or size. We've shipped enough bots to know which steps are non-negotiable. Discovery, strategy, backtest, build, deploy, optimize.
30-60 minute call. We listen, we clarify, we write a one-page spec. The spec is yours regardless of next steps.
Output: 1-page technical brief - rules, risk envelope, kill-switch conditions, infra approach, rough estimate.
Detailed strategy doc with all edge cases enumerated. You read it, push back on anything that doesn't match your intent, sign off before we write code.
Output: 3-5 page strategy doc with config schema, risk matrix, expected behaviors.
Strategy is implemented against 6-18 months of historical Polymarket / Kalshi data. We tune. We report. If the backtest is materially negative or marginal after costs, we say so and propose changes.
Output: Backtest report - PnL curve, drawdown, hit rate, Sharpe, fee/gas accounting. Honest about every cost.
Typed code (Python or TypeScript), structured logging, telemetry, error handling, ops runbook. Code in your Git org from day one - no surprise handover.
Output: Working bot in your repo. Grafana dashboard. Discord/Telegram alerts. Operational runbook.
Two-week paper trading on live data. Side-by-side comparison with backtest expectations. Live with 10% of target size. Scale up over 2-4 weeks based on observed behavior.
Output: Live bot. Daily summary emails. On-call channel for first 30 days. We refuse to ship without paper trading green.
Optional ongoing retainer for tuning, new strategies, market regime adaptations. Most clients take this on their Pro and Enterprise projects; Starter clients usually self-maintain after handoff.
Output: Monthly review call + ad-hoc adjustments. Quarterly strategy report.
Send us a quote request or book a discovery call. Either path starts the same way: a one-page spec, yours regardless of what comes next.