Quantum ML for the physical world
Small quantum-classical models that do real work on today's hardware.
Mission
Small quantum-classical models that do real work
We build quantum-classical hybrid models that run on today's hardware — not on a roadmap. Our approach is practical, not theoretical: train on real data, publish honest results, and focus on problems people already have.
Cyber defense. Earth observation. Anything where a quantum circuit can find a correlation a classical filter misses. We're building the practical quantum stack — one result at a time.
Research
What we're building
Cyber defense
Hybrid quantum ML for passive OS fingerprinting. Variational quantum classifiers on 20-qubit hardware, competitive with classical baselines using dramatically fewer parameters.
Full results at DEF CON 34
Earth observation
Synthetic Aperture Radar data maps naturally onto small variational circuits. Early-stage research in quantum signal extraction for satellite imagery.
SAR · Quantum signal extraction
Foundations
Gradient stabilization for variational quantum models. Adaptive parameter-shift rules, noise-aware clipping, and dynamic learning rate scheduling.
Gradient silence · Stabilization
Approach
Hybrid
Quantum circuits handle signal extraction. Classical models handle the decision layer. The best of both architectures, combined.
Small on purpose
10-20 qubit circuits that fit on current hardware. Not a 1,000-qubit roadmap — models that work on the processors available today.
Real data
We train on actual network traffic, sensor feeds, and operational datasets. Not synthetic toy problems. Results you can trust.
Publication
Your Packets Are Showing
Hybrid Quantum ML for Passive OS Fingerprinting
Full benchmarks are under wraps until our DEF CON 34 debut. The complete results and paper drop August 2026 — check back after the talk.
Team
The lab
La Alsulaim
Co-founder / CEO
CS, University of Pittsburgh
Jae Sung Kim
Co-founder / CTO
CS, Virginia Tech · MS Quantum, UMD
Plus researchers at Carnegie Mellon University and the University of Pittsburgh.
Our Partners
Open to collaborations, co-authors, and funding partners.