Quantum computing for the institutions that move money.
We build quantum and quantum-inspired software for banks, asset managers and insurers. Optimize portfolios under real-world constraints, speed up risk simulation, and move your encryption to post-quantum standards before it becomes urgent.
What we solve
Six problems where the mathematics favors quantum methods. We're specific about which ones pay off today and which are still maturing.
Portfolio optimization
Search far larger allocation spaces under real constraints — cardinality, turnover, sector and ESG limits — and rebalance more often.
Method: Quantum annealing, QAOA, quantum-inspired solvers
Availability: Available now
Risk simulation
Reach the VaR and CVaR precision you need with fewer scenarios, so intraday risk becomes practical instead of an overnight batch.
Method: Quantum amplitude estimation, GPU Monte Carlo
Availability: In pilots
Derivatives pricing
Price path-dependent and exotic options faster, with error bounds that both your desk and your model validators can check.
Method: Quantum Monte Carlo, tensor networks
Availability: In pilots
Fraud and AML detection
Find subtle patterns in high-dimensional transaction data and send fewer false alerts to your investigators.
Method: Quantum kernel methods, quantum-inspired machine learning
Availability: In research
Post-quantum cryptography
Find every place you rely on RSA and elliptic-curve cryptography, rank what to migrate first, and move to NIST-standard algorithms.
Method: Cryptographic inventory, ML-KEM, ML-DSA
Availability: Available now
Quantum-inspired analytics
Get value on classical hardware today with quantum-inspired algorithms on GPUs, for credit scoring and scenario generation.
Method: Tensor networks, simulated bifurcation
Availability: Available now
Interference, put to work
Banknotes resist forgery with interference: fine engraved lines that only line up one way. Quantum computers use the same physics. Waves of probability reinforce the right answers and cancel out the wrong ones.
For Monte Carlo methods, the engine behind VaR, CVaR and option pricing, that changes how fast error shrinks. Classical sampling has to quadruple its samples to halve the error. Quantum amplitude estimation only has to double its queries.
- Classical Monte Carlo
- 1 million samples
- Quantum amplitude estimation
- 1,000 queries
1,000× fewer evaluations
Show as a table
| Target precision | Classical Monte Carlo | Quantum amplitude estimation | Difference |
|---|---|---|---|
| ±1% | 10,000 samples | 100 queries | 100× |
| ±0.5% | 40,000 samples | 200 queries | 200× |
| ±0.1% | 1 million samples | 1,000 queries | 1,000× |
| ±0.05% | 4 million samples | 2,000 queries | 2,000× |
| ±0.01% | 100 million samples | 10,000 queries | 10,000× |
| ±0.005% | 400 million samples | 20,000 queries | 20,000× |
| ±0.001% | 10 billion samples | 100,000 queries | 100,000× |
Idealized scaling with constant factors set to one. Real-world gains depend on circuit depth, error correction and data loading, so we benchmark on your workloads before you commit.
How a pilot works
Twelve weeks, one problem that matters to you, and a clear answer at the end.
Scope
Weeks 1–2
We pick one high-value problem with your team and agree on the benchmark to beat.
Formulate
Weeks 3–5
We translate it into a quantum-ready form and connect to your data inside your own cloud.
Benchmark
Weeks 6–10
We run it on GPUs, quantum-inspired solvers and quantum hardware, and compare every result with your current approach.
Decide
Weeks 11–12
You get the results, the code and a production plan, or a plain answer that it isn't worth it yet.
We run on superconducting, trapped-ion, neutral-atom, photonic and annealing hardware, plus GPU simulation, and we aren't tied to any single provider.
Quantum-safe before it's urgent
Encrypted data stolen today can be decrypted once large quantum computers exist. Financial records stay sensitive for decades, so the migration has to start now.
- 2024
NIST publishes the first post-quantum standards: ML-KEM, ML-DSA and SLH-DSA.
- 2025
NIST selects HQC as a backup algorithm for key exchange.
- 2030
NIST's draft transition plan deprecates 112-bit RSA and elliptic-curve algorithms after this year.
- 2035
The same draft plan disallows them entirely after this year.
How we protect your data
- Runs in your cloud or data center
- Deploy inside your own AWS, Azure or Google Cloud account, or on premises.
- Post-quantum encryption
- Data in transit and at rest is protected with ML-KEM and ML-DSA, the NIST standards.
- Only math leaves your walls
- Quantum hardware receives abstract problem formulations, never client names, trades or balances.
- Audit-ready by default
- Every job is logged with its inputs, versions and results for model risk review.
Questions from risk and technology teams
Do we need our own quantum computer?
No. We reach quantum hardware through cloud providers and run quantum-inspired methods on classical GPUs. Your team works with us through an API and the tools it already uses.
Is quantum advantage real today?
In part. Quantum-inspired methods already beat standard approaches on some problems using classical hardware. Large-scale advantage on quantum hardware is still arriving. We benchmark both on your data and tell you plainly which one pays off.
What happens to our data?
The platform runs inside your cloud account or data center. Only abstract mathematical formulations are sent to quantum hardware, never client names, trades or balances.
Why start the post-quantum migration now?
Attackers can record encrypted traffic today and decrypt it later. Replacing cryptography across a bank's systems takes years, and NIST's draft plan phases out RSA and elliptic-curve algorithms between 2030 and 2035.
What does a pilot cost?
Pilots are fixed-scope and fixed-fee, priced by the size of the problem. Tell us what you'd like to solve and we'll send you a proposal.
Start with one problem
Tell us what you're working on. We'll reply within two business days with a proposed pilot.
Prefer email? pem@skewbits.com