Results

A worked example

Scenario: Comparing classical Markowitz portfolio optimization against a quantum computing approach for a four-asset portfolio.

Using mock data for four assets with expected returns of 8%, 12%, 10%, and 15%, and a correlation structure ranging from 0.1 to 0.4, the classical mean-variance optimizer finds the minimum-variance portfolio and traces the full efficient frontier, the curve of best possible risk and return combinations.

The same problem, framed as a QUBO (Quadratic Unconstrained Binary Optimization) and solved via QAOA on quantum hardware, searches that same space through a fundamentally different computational approach. The scaffolding for that comparison is already built (see quant/portfolio_optimization.py); running it live against real IBM Quantum hardware is the next phase of this project, detailed further in Future Improvements.

A personal finance example

Scenario: Retirement contribution realism check.

A user entering a monthly contribution of $32,000, intending it as an annual figure matching their 401k limit, would previously have received a technically correct but wildly misleading projection with no explanation offered. The retirement calculator now asks for the annual figure directly, matching how contribution limits are actually quoted, checks it against the real limit configured in settings.json, and flags anything that looks like a unit mistake, all while still running the calculation exactly as entered.

This kind of input validation and unit clarity work is not glamorous, but it is precisely the difference between a working demo and software someone could actually trust with their own numbers.

More worked examples and screenshots will be added here as the project develops further.