Confidential - VaultSync Solutions Inc. - Authorized Investors & Partners Only

GDBS & CORA

Comprehensive briefing and talking points - what they are, what they do, the design decisions behind them, their development, and the roadmap.

Briefing 2026-05-19 Investor / partner access

Summary and Positioning

GDBS (Geometric Database System) is a precision-computing substrate and delivery platform. It executes scientific and engineering computations with a number system whose error is a first-class, inspectable quantity rather than hidden machine rounding, and it delivers that capability through a browser-accessible application rather than a cluster queue.

CORA (Conversational Retrieval Agent) is a retrieval-deterministic interface built on the GMDBS geometric substrate. Every answer it returns is composed from specific corpus entries at known positions on a toroidal manifold. The same query against the same corpus returns the same answer, and every sentence is traceable to its source. CORA does not generate language probabilistically; it navigates a geometric layout.

GDBS

The precision and verification platform: GeoNum arithmetic, the physics and engineering modules, and the HPC-grade validation suite.

CORA

The geometric retrieval product: GMDBS addressing plus the deterministic retrieval and composition pipeline, hosted at cora.getvaultsync.com.

The GDBS offering foregrounds precision and verification; GMDBS retrieval is kept minimal inside GDBS and is the basis of CORA as a separate product.

Core philosophy. Determinism, transparency, and verifiability. Numerical error is reported, not concealed. Retrieval answers are bounded by a corpus and auditable, not synthesised. Results are reproducible and, where the substrate permits, available without network access or accelerator hardware.

Legal entity. VaultSync Solutions Inc., a Wyoming C corporation.

GDBS - Geometric Database System

2.1 Definition and intent

GDBS computes physical and engineering quantities at HPC-grade precision with explicit uncertainty tracking, delivered in a standard browser. It is not positioned against the raw throughput of petascale or exascale clusters; it is positioned against the access model of those clusters - instant availability, no scheduling queue, transparent uncertainty, and reproducible results.

2.2 Substrate architecture

The three-layer split is deliberate: Rust for substrate and engine code, .NET 8 for the API, HTML/JS for the frontend. Engine code is never written in the API or frontend languages.

2.3 GeoNum precision system

Transparent uncertainty API. Every value reports getUncertainty() and getRelativeUncertainty(). One GeoNum class; domain specialisation changes only the zone-boundary configuration (Appendix D). Precision tiers: 256 / 512 / 1024 / 2048 shades.

2.4 GMDBS geometric model

Compression. Roughly 2 GB to 448 MB (about 78 percent) at a representative tier; per-record about 32 bytes to about 7 bytes; multi-node CPU speedup near 9.7x.

2.5 Web platform and product tiers

Standard

Single-physics exploratory tools, included with an authenticated account.

Pro / Advanced

Coupled multi-physics, batch sweeps, optimiser and predictor, validation suites.

HPC Lab

Full research sandbox with raw engine access and custom workflows.

Pricing is tier-based; authoritative figures are those published on the live platform (Appendix H). A WebGPU compute path with automatic CPU execution and a GPU-capable charting layer (opt-in per module) round out the platform.

2.6 Module catalogue

Domain modules: Plasma, Fluids, Quantum, Materials / Geometric MD, Topology, Geophysics, Medical / Molecular, Cosmic, Theory. HPC Lab research engines: Multigrid Poisson, FDTD electromagnetics, Phase-Field (Cahn-Hilliard / Allen-Cahn), Compressible Euler (HLLC + MUSCL), DMRG quantum many-body. Substrate engines include BSSN/Z4c numerical relativity, a LIGO pipeline, three-body dynamics, manifold prediction, finance, MLE, RF compaction, multi-hypothesis voting, DFT kernels, and an arXiv parser; an HPC-IO group and the Adv multi-physics module complete the platform (full list in Appendix A).

2.7 Validation and results

CORA - Conversational Retrieval Agent

3.1 Definition and intent

CORA is retrieval-deterministic. There is no generative model in the answer path. Every word is composed from specific corpus seeds at known positions on the toroidal manifold, so the same query against the same corpus produces the same answer with auditable provenance. When the answer is not in the corpus, CORA says so; out-of-corpus invention is structurally impossible. Answers are located, not produced - the geometric layout is itself the relevance function.

3.2 How retrieval works

  1. Plan-cache lookup - a hashed query may return at sub-microsecond latency from a bounded plan cache (~2048 entries).
  2. Depth and tier selection - adaptive performance tier from rolling stats; self-referential queries held to a higher acceptance threshold.
  3. Integrity gate - blocks extraction, identity manipulation, harmful requests.
  4. Routing - normalised content words; first routable word selects one of 24 ontology hubs and a target sub-toroid.
  5. Crawl and characterise - candidates scored by a weighted geometric relevance function (curated prior, Poincare geodesic distance, topic-hash, hub specificity, word overlap, bloom mass, usage reinforcement).
  6. Divergent retry - wider re-run if the best score is weak and budget allows.
  7. Linguistic rerank - content words as a hard filter; coverage and modifier rescale; code spans excluded.
  8. Composition - retrieved (dominant seed) or constructed (fragments weighted by retrieval score); no learned parameters.
  9. Fidelity gate - checked against source for entity coverage and overlap.
  10. Traceability verification - traceable fraction always measured.
  11. Assembly - response plus a footer of mode, confidence, traceability, routed hub and sub-toroid.
  12. Reinforcement and cache insert.

Session state (turn counter, accumulated ontology positions, interaction profile, conversation context) calibrates response style but does not introduce non-determinism into what is retrieved. Every response exposes confidence, traceability, mode, routed hub and sub-toroid in an inspectable trace panel.

3.3 Deployment modes

Browser-only (GMDBS WASM)

~413 KB binary, ~650 MB seed bundle (~2.17M seeds), ~850 MB wasm memory, 0-4 ms query, zero network after load - air-gap capable.

Server-hosted (.NET + Wasmtime)

gdbsio driver, memory-mapped shards, 4.5M+ seeds, sub-100 ms boot, sub-millisecond cached query.

3.4 Hosting and integration

Hosted at cora.getvaultsync.com. Inside the GDBS application a thin bridge calls status and query endpoints; the element appears only after an appropriate tier is confirmed, a computation is registered, and the status endpoint responds, and is hidden silently otherwise.

3.5 Relationship to GMDBS and trajectory

CORA is the productised expression of the GMDBS substrate. The longer-term trajectory treats the system as a geometric brain in which retrieval is emergent - a structured navigable network whose deformation, not a re-ranking heuristic, changes what is recalled.

Design Decisions and Rationale

Development History

Roadmap

Talking Points (condensed)

Precision

Error is a first-class, inspectable quantity. HPC-grade multi-scale precision in a browser. 0.27 percent across ~60 orders of magnitude at 0.345 shade of drift.

Transparency

Uncertainty queryable on every value; drift reports cancellation cost; CORA answers fully traceable.

Determinism

Same input, same output. CORA says "I do not know" rather than inventing.

Accessibility

No queue, no scheduler, no cluster account; a browser and air-gap client.

Geometric model

Position is the key (HVP); pentagon scaffold, Lucas-gap tiers, ~78 percent compression.

Validation

QED c1-c5 inside published bars; convergence 3.946; cosmic calibrated to 14 references.

Product structure

GDBS for precision/verification; CORA for deterministic retrieval; tiered access.

Entity

VaultSync Solutions Inc., Wyoming C corporation.

Glossary

BSSN / Z4c
Formulations of the Einstein equations used by the numerical-relativity engine.
Bloom-anchored seed
A corpus entry with a compact 256-bit bloom signature for fast candidate filtering.
Closed-Form Validator
The HPC Lab suite reproducing QED coefficients c1-c5 against published references.
CORA
Conversational Retrieval Agent: the retrieval-deterministic product built on GMDBS.
Drift compartment
The first-class accumulated-error field on every GeoNum value.
DSO
Dynamic Scalar Operator: the framework whose single free parameter is g-dagger.
GDBS
Geometric Database System: the precision-computing substrate and browser platform.
GDBSIO
The WebAssembly substrate runtime loaded by the .NET host.
GeoNum
The base number system: zones, 2048 shades, first-class drift, transparent uncertainty.
GMDBS
The geometric mesh and addressing model: toroidal pentagon mesh, HVP, Lucas-gap tiers.
gms
The compressed content-addressed store (lz4 + 64-bit xxhash dedup) behind gdbsio.
g-dagger
DSO threshold constant, 1.2 x 10^-10 m/s^2; the single free parameter.
HVP addressing
The (H, V, P) coordinate scheme where manifold position is the key.
Lucas tier
Level-of-detail tiers whose sizes follow the Lucas sequence.
MZV
Multiple zeta value: the depth-2 transcendental basis in the QED cascade closure.
Pentagon scaffold
The distributed five-vertex anchor structure; a scaffold, not the geometry.
Poincare disk
The local hyperbolic geometry per pentagon; geodesic distance contributes to scoring.
Retrieval-deterministic
Answers composed from located seeds, reproducible and traceable, no generative step.
Seed
A corpus entry positioned on the manifold; the unit CORA retrieves and composes from.
Shade
One of 2048 micro-precision subdivisions within a zone.
SUMO ontology
The tile-to-hub-to-domain overlay (24 hubs) routing a query to a mesh region.
Trust tier
Reliability classification derived from drift relative to the 1.0-shade gate.
VaultSync Solutions Inc.
The legal entity (Wyoming C corporation) behind GDBS and CORA.
Zone configuration
The domain-specific set of zone boundaries; the only tunable distinguishing domains.

Reference Tables

Appendix A. Module catalogue

ModuleWhat it computes
PlasmaIdeal-MHD stability, equilibrium, figures of merit for tokamaks, stellarators, FRCs
FluidsReynolds, boundary layer, compressible flow, lattice-Boltzmann, FDTD, heat transfer
QuantumGround-state DFT, multi-electron atoms, molecular benchmarks, circuit tools
Materials / Geometric MDMolecular dynamics, elastic and thermal properties, phase diagrams, ablation
TopologyFaddeev-Skyrme hopfions, 1D chaos, SU(2) structure from the QED cascade
GeophysicsGravitational fields, seismic propagation, stress, heat flow, earthquakes
Medical / MolecularBinding, protein stability, nanoparticle design, QSAR, SASA, virtual PCR
CosmicMatter power spectrum, halo mass function, weak lensing, CMB, galaxy rotation
TheoryPhysical and cosmological constants, black-hole thermodynamics, DSO framework
HPC Lab enginesMultigrid Poisson, FDTD, Phase-Field, Compressible Euler (HLLC+MUSCL), DMRG
Substrate enginesBSSN/Z4c, LIGO pipeline, three-body, manifold prediction, finance, MLE, RF compaction, multi-hypothesis voting, DFT, arXiv parser
HPC-IO / AdvForm generator, report builder, arXiv solver, CSS design system; multi-physics coupling and batch sweeps

Appendix B. Validation results

ResultObservedReference / target
Hawking radiation relative error0.27 %analytical
Hawking radiation drift0.345 shade< 1.0 shade gate
BSSN factor-2 self-convergence3.9464.0 (fourth order)
BSSN wall-time scaling slope3.0483.0 (cubic)
BSSN billing calibration~23.94 us/cell/stepmeter calibration
Bounded evolution (pure BSSN)~5 light crossingsformulation limit
Bounded evolution (Z4c)7 light crossingsnext-phase: 1000
DSO m_p / m_e prediction18360.0076 % from CODATA
Cosmic calibration references14 canonicalpublished values

Appendix C. QED cascade coefficients

CoefficientReferenceResidual vs referenceAcceptance
c1Schwinger 1948 (1/2 exact)~5.6e-17floating-point
c2Petermann 1957 / Sommerfield 1958~1.3e-14floating-point
c3Laporta-Remiddi 1996~8.5e-13floating-point
c4Aoyama et al. (4-loop)~4.0e-11inside published bar
c5Kinoshita 2017 (5-loop)~5.7e-10inside published bar

Transcendental basis: zeta(2), zeta(3) by Apery series; zeta(5,7,9) by Euler-Maclaurin; depth-2 multiple zeta values; genus-1 Eisenstein lattice sums. Reproduced via the release test suite.

Appendix D. GeoNum zone configurations

ConfigurationBoundary styleRange
TheoryLucas-sequence logarithmic10^-35 to 10^30
QuantumTight eV-scale~10^-2 to 10^2 eV
FluidsUniform gridgrid-dependent
PlasmaPlasma-frequency aligned10^3 to 10^12 Hz
MaterialsLattice-symmetric10^-10 to 10^-8 m
GeophysicsSpherical-harmonic aligned10^3 to 10^7 m
BallisticsMach-number scaled10^-1 to 10^4 m/s
Wide uniform1-unit log spacing~10^-110 to 10^110

Precision tiers (shade count): 256 exploration, 512 single-domain production, 1024 multi-scale, 2048 cross-decade and publication.

Appendix E. Technology stack

Appendix F. Product boundary and SBIR scope

GDBS and CORA are separate products on a shared substrate. GDBS scope is GeoNum precision and verification plus the physics/engineering modules; GMDBS retrieval is minimal inside GDBS. CORA is the GMDBS geometric-retrieval product. SBIR Phase I is scoped to the GDBS side (four tasks, production-port and minimal-hosting framing); CORA's retrieval substrate is not foregrounded in the GDBS SBIR scope.

Appendix G. Key constants

Appendix H. Pricing and tier note

Pricing is structured by tier: Standard (included with an authenticated account), Pro / Advanced multi-physics, and HPC Lab. Published source documents and prior internal guidance differ on specific dollar figures; this briefing states the tier structure and defers exact pricing to the live platform. Quote the tier name and confirm the current amount from the live site rather than a fixed number from memory or older documents.