Qantara
Co-built a hybrid quantum-classical portfolio optimizer balancing return, risk, diversification, and ESG priorities.
- Field
- Quantum computing
- Role
- Quantum optimization engineer
- Recognition
- Challenge winner · GenQ Hackathon 3rd overall
The challenge
Qantara had to turn competing goals, including return, risk, diversification, portfolio size, and ESG exposure, into an optimization problem that could be run and explained during a hackathon.
Kean Louis Rosales and I built Qantara during the 2025 GenQ Hackathon in Singapore for QCentroid’s sustainable portfolio optimization challenge.
My contribution
Kean and I worked across the optimization pipeline and supporting visualizations. We implemented the CVXPY baseline, QUBO formulation, and QAOA experiments in Python using pytket and Qiskit Aer, with a Quantinuum-compatible execution path.
The hybrid workflow
The system prepared price and ESG-oriented inputs, including simulated ESG histories, then used classical filtering to narrow the candidate set. It translated the binary asset-selection problem into a QUBO, ran QAOA experiments through local simulation and Quantinuum-oriented tooling, and retained a CVXPY mean-variance baseline.
The formulation balanced several objectives:
- expected return;
- risk;
- diversification;
- ESG exposure;
- portfolio-size and selection constraints.
The comparison
The two formulations answered different questions. The QUBO selected assets as binary variables, while the CVXPY model optimized continuous portfolio weights. The hackathon evaluation compared them through a shared workflow, making the difference between asset selection and portfolio weighting explicit.
The result
We won the Portfolio Optimization challenge presented by QCentroid and placed third overall at the Singapore GenQ Hackathon.