A PDE pricing engine that doesn’t time-step. Arithmetic that’s bit-identical across every desk. Single-solve uncertainty quantification for the overnight risk run. And an evidence base built to survive model validation.
Proofs of concept
Each POC opens with a scenario, walks through what LRDE changes, and ends with a benchmark number plus a SHA-256 you can re-derive on your own hardware.
Single vanilla call on a BS MOL grid, 9 configurations. LRDE is 5–7× faster at matching precision.
Full strike × maturity grid up to 6,000 cells. LRDE is 26–41× faster than scipy BDF.
25-trade portfolio × 200 stochastic scenarios. 8.5–12× speedup, 29 micro-units agreement per trade PV.
Finance capabilities
The same deterministic solver, evidence-based tooling, and audit-grade research platform show up across derivatives pricing, overnight risk, card authorization, reconciliation, model risk, and regulatory capital.
Derivatives Pricing
Every options library reduces to integrating a linear PDE in time - and every library time-steps it. SolvLRDE computes the answer as a single contour integral in the Laplace domain, independent of maturity and stiffness. One LU factorization, reused across every strike, every maturity, every scenario in the same vol bucket. 5–40× faster than scipy BDF on benchmark workloads, with sub-basis-point agreement.
Overnight Risk & xVA
Overnight portfolio valuation, VaR, Expected Shortfall, and xVA fan-outs all bottleneck on the same inner loop: re-price every instrument under every scenario. SolvLRDE’s LU amortization across scenarios makes the fan-out structurally cheaper than any time-stepper. Paired with SolvSRK-UQ’s single-solve uncertainty quantification, the entire risk run - prices and confidence intervals - lands before the opening bell.
Card Authorization & Payments
Card networks compress a rich upstream picture into twenty standardized fields before the issuer decides. Moridain sits on the network, extracts ten bounded intelligence signals from the full picture, and delivers them alongside the standard message. Banks running rules engines outperform banks running machine learning - with a tenth of the regulatory exposure.
Cross-Desk Reconciliation
The front-office pricer, the risk engine, the back-office settlement system, and the regulatory report all compute on different hardware and produce slightly different numbers. SolvNum is a deterministic real-number primitive that produces bit-identical arithmetic across x86, ARM, GPU, and WebAssembly. One SHA-256 receipt proves every desk agrees.
Model Risk & Audit
Model risk management requires knowing why a particular solver was chosen and whether it is appropriate for the system it solves. SolvScout fingerprints your ODE/PDE system and benchmarks every candidate solver against it - dead zones called out, not hidden. SolvJump study runs and SolvSRK proof bundles archive the decision for replay. “Why this solver?” has a defensible answer at model validation.
Regulatory Infrastructure
Regulatory capital calculations under FRTB IMA require reproducible risk numbers across systems and time. SolvNum’s cross-platform determinism means the approved model, the production run, and the audit replay produce the same bits. SolvJump and account release bundles provide the replayable evidence trail. The combination gives model-risk and compliance teams a turnkey argument for regulatory acceptance.
Want to see the numbers on your own book?
We offer a bounded-scope read-only benchmark against a sanitized snapshot of your real vol surface and trade book. Two weeks, $25K, fully credited toward a Year-1 license.
The product line
Each product brief below opens with a scenario your team will recognize, then walks through what it is, who uses it, why it’s defensible, and how the claims survive model validation.
Answer at time T without paying for every micro-step in between.
Long horizons and stiff linear structure punish time-marching solvers — you pay in steps, in stiffness, and in repeated factorizations across scenarios. SolvLRDE targets the linear response directly so scenario grids, strike ladders, and EMTP-style studies scale with factorization reuse, not with horizon length.
Uncertainty bands without a Monte Carlo farm.
Certification, risk, and autonomy teams need error envelopes — but thousand-run Monte Carlo grids miss program milestones. SolvSRK-UQ delivers calibrated uncertainty alongside the trajectory in one pass, fast enough for control loops, overnight risk batches, and evidence packages that auditors can reproduce.
Plan, flight, and audit replay agree — provably.
When three teams run the same mission math on three platforms and get three different answers, the argument is never about physics — it is about arithmetic. SolvNum gives you bit-identical reals across CPU, GPU, and WASM with a cryptographic receipt, so V&V, coalition replay, and capital models stop fighting floating-point drift.
Decisions get context the wire was never designed to carry.
Standard messages throw away the upstream picture before your rules engine or fusion node sees it. Moridain puts bounded intelligence back on the rail — so fraud, fusion, and operator queues rank on signal, not on whoever shouted loudest in the compressed feed.
Stop guessing which solver your model actually needs.
The wrong integrator costs months of false confidence before a stiff cell or noise case fails in review. SolvScout fingerprints your system, benchmarks every serious candidate on your problem, and shows a ranked table — including when a competitor wins — so model validation starts from evidence, not habit.
Shortlist integration candidates in minutes; confirm in hours—not weeks of sweeps.
Integration programs bleed calendar on hand-rolled parameter marathons and configs that only survive until the next variant. SolvJump ranks plausible integrator settings in minutes on the bench, then drives SolvSRK proof runs on finalists so you sign off in hours—not days or weeks babysitting sweeps.
Engagement model
Read-only benchmark on your book in two weeks. Shadow run alongside your existing pricer for one month. Production pilot on one desk, one workload. Every option credits in full toward a Year-1 license. Every numerical claim traces back to a test ID, a seed, and a verdict in a live evidence register.
Governed research, licensable IP, partner programs.
Company brief, evidence, and data room access.