Quantum machine learning has been overhyped and under-delivered in roughly equal measure. This week is not a sales pitch for either quantum or topology. It is a working week for people who want to know where, specifically, these tools give measurable lift over strong classical baselines on financial problems. Sometimes they do. Sometimes they do not. We are interested in which.
The bootcamp covers variational quantum circuits and quantum-kernel methods at the level needed to use them (Qiskit, PennyLane), topological features for non-stationary financial time series, and the open question of whether the two together earn their compute budget. Half the week is lectures and labs; half is a research workshop where participants and faculty work on shared problems.
Program Overview
One intensive week at AIRINA Labs, Cotonou — with a hybrid option for remote workshop participants. Mornings are lectures, afternoons are hands-on labs in Qiskit and PennyLane, and the final day doubles as a research workshop where participants and faculty work in pairs on open questions drawn from the curated list. The cohort leaves with a shared draft research note.
Lead instruction by AIRINA's quantum group; co-instructors in quantum computing and quantitative finance drawn from AIRINA's network of academic and industry collaborators (named four weeks before the cohort).
Program structure
- Days 1–4 — lectures and labs. Variational circuits, quantum kernels, topological features for financial time series, hybrid pipelines, case studies on real datasets.
- Day 5 — research workshop. Pair work on a chosen problem from the open-questions list, afternoon presentations, and a shared group writeup.
- Cohort size. ~15 participants. Application-based.
Certificate
Successful completion of the pair-written research note (5–8 pages) earns an AIRINA Quantum-Topological Methods certificate — graded final. Honest negative results are welcome; assessment is on the quality of the question and the rigor of the comparison, not on whether the quantum or topological method "won."
Learning Outcomes
By the end of the program, participants will be able to:
- Build a variational quantum classifier in PennyLane or Qiskit, train it, and benchmark it honestly against a strong classical baseline.
- Compute quantum kernels for a real financial dataset and identify the regime where they outperform RBF / Matérn / random-features baselines.
- Apply topological features (persistence images, Betti curves) to financial time series for regime detection, with proper attention to non-stationarity.
- Read a recent QML or topological-finance preprint, identify its claim, find its weakest baseline, and decide whether the claim survives.
- Recognize the cases where neither quantum nor topology helps — and say so clearly when they do not.
Program curriculum
Qubits, gates, circuits in PennyLane and Qiskit. Variational circuits, parameter-shift rule, QAOA on a toy portfolio problem. Quantum kernels and what they correspond to classically.
Cubical and Vietoris–Rips filtrations applied to log-return series. Persistence images and Betti curves as features. Detecting regime change topologically. Why the stability theorem matters when the data is non-stationary.
Quantum-enhanced feature spaces seeded with topological features. Hybrid quantum-classical pipelines. The open question: when does compute spent on the quantum side beat the same compute spent on classical deep learning?
Volatility-surface modelling, regime detection in FX, portfolio construction under heavy-tailed correlation structure, credit-risk on small-sample data. Each case study with its honest baseline and a clear verdict.
Participants work in pairs on a chosen problem from the week's open-questions list. Afternoon: presentations and group writeup. The cohort leaves with a shared draft research note.
Who Should Attend
This bootcamp is for quants, advanced graduate students, and researchers who want a working — and sober — view of what quantum and topological tools contribute at the frontier of quantitative finance.
- Quants and financial-engineering practitioners at banks, hedge funds, asset managers, or fintechs.
- Advanced graduate students (PhD or final-year master's) in mathematics, physics, or computer science with a finance interest.
- Researchers in quantum machine learning curious about the topology side — and vice versa.
Prerequisites
- Quantitative finance. Comfort with stochastic calculus basics, time-series modelling, and portfolio optimization.
- Mathematics. Linear algebra at graduate level. Some prior exposure to quantum computing helps; not strictly required if you can move fast in week 1.
- Programming. Working Python; experience with Qiskit or PennyLane is a plus.
Selection
Cohort fills slowly to keep the workshop tight. Applications open early 2027. We ask for a CV and a short paragraph on what you would use this for.
Brochure
The detailed program brochure (PDF, EN/FR) is sent on request — including day-by-day curriculum, faculty profiles, the open-questions list, and the cohort calendar.
To receive the current brochure, write to contact@airina.africa with "Quantum-Topological Methods — brochure request" in the subject. The brochure is updated each cohort; we send the version current at the time of your request.