Build a Hybrid Quantum AI Solution
Objective: Design, implement, and present a hybrid Quantum–Classical AI model that addresses a real-world, ethically-scoped problem. Deliverables include a runnable notebook, project report, and a short presentation. Successful completion awards the Quantum AI Specialist Certificate.
Estimated effort: 6–8 weeks (part-time)
Prerequisites: Completion of Modules 1–8 (foundational quantum computing, Quantum AI patterns, ethics, and hands-on labs).
What learners will accomplish
- Build a full hybrid pipeline (data → encoding → quantum subroutine → classical model → evaluation).
- Demonstrate practical choices for quantum components (PQC/VQE/QAOA/quantum kernels) and justify trade-offs.
- Include an ethical risk assessment and deployment checklist.
- Produce reproducible code and a professional presentation—suitable for portfolio or employer review.
Capstone timeline & milestones (6 weeks)
- Week 0 — Topic selection & team formation (optional)
- Week 1 — Project proposal (due Day 7)
- 1-page problem statement, goals, dataset, chosen quantum approach, evaluation metrics.
- Weeks 2–3 — Development sprint 1
- Data processing, baseline classical model, prototype quantum module.
- Week 4 — Development sprint 2
- Integration, hyperparameter tuning, and initial evaluation.
- Week 5 — Ethics & scalability review
- Complete the ethics checklist, run lightweight scalability tests, document limitations.
- Week 6 — Finalize and present
- Final report (≤5 pages), runnable notebook, 5–10 slide presentation, optional demo video.
Milestone checkpoints: weekly mentor feedback and one formal mid-project review at end of Week 3.
Deliverables
- Project proposal (PDF/Markdown) — problem, goals, data sources, methods, and risks (max 1 page).
- Runnable Jupyter notebook — clear install/run instructions and reproducible outputs.
- Final report (≤5 pages) — background, methods, results, limitations, ethics & deployment checklist.
- Presentation (5–10 slides) — concise demo-ready deck.
- Optional: 3–5 minute demo video or recorded walk-through.
Submission format: GitHub repo or ZIP with notebook(s), report, slides, and a README containing reproduction steps.
Assessment & Grading Rubric
| Criterion | Weight | Description |
|---|---|---|
| Technical correctness & reproducibility | 40% | Code runs end-to-end; results are replicable. |
| Empirical evaluation & analysis | 25% | Sound metrics, baselines, ablation studies, and visualization. |
| Ethical & safety analysis | 15% | Clear risk assessment, mitigation plan, and compliance considerations. |
| Novelty & design choices | 10% | Innovative or well-justified hybrid approach. |
| Presentation & documentation | 10% | Clear report, readable notebook, and concise slides. |
Pass threshold: 70% overall and no critical ethics violations.
Resubmissions: Allowed once after mentor feedback for up to 2 weeks.
Recommended project archetypes & starter ideas
- Quantum Drug Screening (Healthcare): Hybrid QSAR + VQE features to rank candidate molecules (small dataset).
- Portfolio Optimizer (Finance): QUBO-based selection solved by QAOA or simulated annealing; compare to classical mean-variance.
- Quantum NLP (Text): Use a PQC feature-transformer with sentence embeddings for classification or semantic search.
- Logistics (Routing): Small VRP/TSP mapped to QUBO with hybrid decomposition and classical post-processing.
- Climate Sampling: Quantum-assisted sampling for a toy stochastic model to estimate uncertainty in forecasts.
Mentorship & support structure
- Mentor allocation: One mentor per project (or team) for weekly reviews.
- Office hours: Twice-weekly drop-in sessions for troubleshooting and guidance.
- Peer review: Mid-project peer feedback session; optional code review swap.
Ethics, IP & Publication policy
- Ethics: Projects must include an ethics assessment describing dataset consent, potential harms, and mitigation measures. Projects that pose dual-use, biological risk, or security sensitivities will require additional review and may be restricted.
- Intellectual Property (IP): Students retain IP by default, but the program may request a royalty-free license to showcase exemplary projects. If collaborators include external partners, clarify IP in the proposal.
- Publication: Public dissemination is encouraged but must not violate safety rules; embargo options available for sensitive work subject to review.
Resources & templates
- Project proposal template (Markdown) — include problem, dataset, methods, and ethical note.
- Notebook template — clear header, environment instructions, and reproducible execution cells.
- Report template (LaTeX/Word) — Background, Methods, Results, Ethics, Appendix.
- Evaluation checklist — reproducibility checklist, ethics checklist, and submission checklist.
(Templates will be added to the course repo; mentors will provide links on Day 0.)
Certification & Badge
Certification criteria: Complete all modules, pass capstone evaluation (≥70%), and submit final artifacts to the course portal.
Showcase & Career support
- Top projects spotlight: Selected projects featured on course site and newsletter.
- Hiring pipeline: Option to share top projects with partner employers (with student consent).
- Career guide: How to present quantum AI projects in portfolios and interviews (resume bullets, demo tips).
FAQ for Capstone Applicants
Q: Can I work in a team?
A: Yes — teams up to 3 are allowed. Each member should contribute and be listed with responsibilities.
Q: What compute is provided?
A: Access to cloud simulators and limited quantum hardware credits (if available). Additional paid credits are the student’s responsibility.
Q: How long does grading take?
A: Mentor feedback is weekly; formal grading is completed within 2 weeks of final submission.
Q: What if my project fails technically?
A: You can still pass if your report documents the failure, presents reasonable analysis, and shows learning — reproducibility and rigor matter more than positive results.
Next steps & CTA
- Step 1: Draft and submit your one-page proposal by Day 7.
- Step 2: Join the Capstone kickoff session (calendar invite).
- Step 3: Clone the course repo and start from the Notebook template.
Ready to start? Click [Start Capstone — Submit Proposal] (course portal link) to begin.