Quantum · QAOA · q-routed conductor · /v1/portfolio

Optimal allocation, found not approximated.

One job: take your assets and risk target, return the portfolio weights that maximize risk-adjusted return. Shaped for QAOA and routed through the q-routed conductor — deterministic today, live Qiskit Phase H — the combinatorial step classical mean-variance solvers approximate as the asset count grows.

Problem size

50 assets

Optimized in minutes, not an overnight batch job.

Solver

QAOA

Quantum Approximate Optimization · ibm_fez / boston / pittsburgh.

Every result

Sealed

Dual-signed receipt + WORM-anchored job record.

Classical mean-variance

Approximates above ~30 assets
Re-runs overnight as constraints grow
No tamper-proof record of the run

QAOA · EpochCore

Explores the full combinatorial space
Minutes per optimization, on demand
Sealed job — reproducible at audit

Run it

Optimize a portfolio.

Enter your assets, set a risk target, and run. With an API key it hits the live /v1/portfolio QAOA solver; without one it runs the mean-variance reference locally so you can see the shape of the result.

Portfolio · register + optimizePOST /v1/portfolio/:id/optimize

API key (optional — blank runs local reference)

Risk target (annualized σ)

Assets · ticker · expected annual return · annual volatility

+ Add asset

Reference solver runs locally · live solver needs a key

SDK

Three lines in Python.

from epochcore_app_primitives import Client
with Client(api_key="...") as c:
    pf = c.portfolio.register(name="core-equity", assets=assets, risk_free_rate=0.04)
    opt = c.portfolio.optimize(pf["portfolio_id"], risk_target=0.20, method="qaoa")
    # opt["weights"], opt["sharpe"], opt["sealed"] — dual-signed