Beta interest
Describe a research use case. Public access is not open.
Get in touchHome / Backtest
Test the idea.
Keep the assumptions.
In-house backtesting research · AI skill in development. The public interface below is illustrative; it does not execute a backtest.
Beta · Internal use and course demonstrations
Execution assumptions
Changing this control selects illustrative outputs only. No data or order is sent to a trading system.
Net profit
164,950
Drawdown
−12.4%
Trades
86
Sharpe
0.91
Illustrative metrics and paths. Backtests do not predict future performance.
01 Data
Preserve the information boundary
Record the instrument, data source, available timestamps, cleaning rules and version.
02 Execution
Explain how an order becomes a fill
Document timing, price conditions, fees, slippage and unfilled orders.
03 Validation
Look beyond the largest profit
Examine risk, nearby parameters, independent periods and failures.
AI Skill · In developmentA research record
A research record
an agent can inspect.
The planned skill will connect a structured request to the research system and return assumptions and limitations with the result. Public access, exact capabilities and pricing have not been released.
Proposed request format · not a live API
{
"instrument": "DCE.i",
"execution": "next_open",
"checks": ["cost_stress", "out_of_sample"],
"result": "report_with_assumptions"
}No credentials, real account or execution access are exposed by this website.
Help shape the next version.
Design partners
Discuss team workflows and a scoped pilot. Terms are agreed individually.
Get in touchAgent access
Machine-readable reports and pay-per-use are planned. Payments are not live.
Get in touch