Build your AI portfolio agent.

A portfolio is a set of assets generated by the platform based on your inputs and workflow. Each portfolio includes one monthly rebalance, one walk-forward backtest, and one full performance report, regardless of your plan.

Yes. Cancel anytime from your account settings. You'll keep access until the end of your billing period. No penalties, no hidden fees.

Core models use standard frameworks like mean-variance optimisation and Sharpe Ratio ranking. The advanced model (Pro only) combines multiple alpha signals into a single ranking system designed to identify higher-conviction opportunities.

Yes, annual plans are 20% cheaper than monthly. You're billed once per year.

Pro is currently in development. Join the waitlist to get early access and discounted pricing when it launches.

No. The platform provides data-driven insights, portfolio suggestions, and analytical tools, not personalized financial advice. You remain fully in control of your investment decisions.

Hierarchical Risk Parity (HRP), Mean-Variance Optimization (MVO), and Conditional Value at Risk (CVaR). HRP clusters assets by correlation structure and avoids inverting the covariance matrix, which makes it less sensitive to estimation error than classical mean-variance on noisy universes. Assets can be ranked by market capitalization, Sharpe ratio, twelve-month momentum, or a composite quantitative ranking.

Historical prices, market capitalization, company profiles, and the treasury yields used as the risk-free rate come from Financial Modeling Prep. Returns used for allocation and backtest simulation are computed on dividend-adjusted prices; the cheaper screening and ranking pass runs on unadjusted closes. Assets without adequate price coverage over the evaluation window are dropped at the screening stage rather than carried forward with interpolated history.

Yes. Round-trip transaction costs can be applied in basis points per unit of turnover, so a high-turnover strategy is charged for its turnover instead of being compared against a frictionless benchmark. Backtests are walk-forward: at each rebalance date the same screen, select, and allocate steps re-run using only data available up to that date.

Only if you choose to enable it. Connecting a broker is optional, and a standard connection runs on one-click approval: it prepares the orders and waits, so nothing reaches the market until you approve it. Unattended rebalancing is available only on specific broker setups, is off by default, and can be switched off at any time. Stored broker credentials are encrypted at rest and are never displayed back to you or written to logs.

Generation is blocked until the next billing period or until you upgrade. Portfolios you have already generated stay available — reports, backtests, and monitoring continue to work. Nothing is deleted for hitting a limit.

No. Portfolio generation, walk-forward backtesting, and performance reporting all work without a broker connection. Connecting one adds live position monitoring and optional rebalancing against a real account; it is not required for any of the research workflows.