Skip to main content

Software Products & Codes by industry

Research & scientific AI Software & Codes

Develop and test reproducible methods for model diagnostics, information theory, and scientific learning. Explore matrix and tensor methods and quantum-assurance evidence across disciplines. Browse software and scientific codes relevant to this area, with methods, screenshots, and resource links.

Browse by industry

Choose an industry to explore relevant software and codes. Research & scientific AI brings together methods used across industries.

10 of 10 entries

Browse entries by name
EnviCloud demos organized by workflow stage, tool family, and domain.
Platforms & applicationsCloud platform

EnviCloud

  • Brings geospatial data, scientific models, and AI analyses into a shared browser workspace.
  • Provides public demonstrations spanning geothermal systems, groundwater, information theory, and resource decisions.
  • Connects maps and interactive diagnostics with reproducible model and analysis workflows.
SmartTensors physics-informed machine learning and R&D 100 recognition.
Scientific frameworksOpen-source framework

SmartTensors

  • Provides a family of unsupervised, supervised, and science-informed machine-learning methods.
  • Uses matrix and tensor decompositions to extract interpretable patterns from multidimensional scientific data.
  • Includes public code and examples for feature extraction, source separation, and large scientific workflows.
TensIT information-retention and computational-cost comparison across candidate tensor resolutions.
Scientific frameworksInformation-theory software

TensIT

  • Compares data representations and resolutions using retained information, structure, occupancy, and computational cost.
  • Investigates nonlinear dependence, redundant attributes, temporal lags, and information that emerges from combinations of variables.
  • Keeps uncertainty, stability, and statistical dependence separate from claims of physical causation.
Q-Sure assurance workflow in dark mode, displaying frozen offline-emulator evidence.
Scientific frameworksQuantum-assurance research

Q-Sure

  • Evaluates evidence, resource budgets, and uncertainty in quantum machine-learning workflows.
  • Uses calibration and explicit reliability gates to support run, measure-more, reroute, or abstain decisions.
  • Compares frozen simulator studies with classical controls without claiming demonstrated hardware or computational quantum advantage.
MADS: Model Analysis and Decision Support.
Scientific frameworksOpen-source framework

MADS

  • Connects model calibration, sensitivity analysis, uncertainty quantification, and decision support.
  • Can analyze numerical, analytical, and machine-learning models through a common scientific workflow.
  • Provides public Julia code, documentation, and examples, alongside the historical C implementation.
Nonnegative matrix factorization with k-means clustering and physics constraints.
Matrix & tensor learningPublic scientific code

NMFk.jl

  • Extracts nonnegative latent signals and their mixing coefficients from matrix data.
  • Combines repeated factorizations, customized k-means clustering, and constraints to examine candidate signal counts.
  • Supports applications such as source separation, pattern discovery, gap reconstruction, and scientific surrogate modeling.
Tucker decomposition into a core tensor and three factor matrices.
Matrix & tensor learningPublic scientific code

NTFk.jl

  • Extracts latent factors from multidimensional arrays using nonnegative tensor factorization.
  • Supports CP and Tucker models, with optional sparsity and physics-informed constraints.
  • Examines repeated solutions and clustering stability to estimate latent feature counts along tensor dimensions.
Three-dimensional subsurface model with geological layers and wells.
Earth-science modelsPublic scientific code

FEHM.jl

  • Provides a Julia interface to the separate FEHM flow-and-transport simulator.
  • Connects simulations of single- and multiphase flow in fractured and unfractured porous media.
  • Serves as an interface module in the MADS and ZEM frameworks; it is not the FEHM solver itself.
Related MADS parameter-calibration and model-diagnostics workflow in dark mode.
Inference & uncertaintyPublic scientific code

BIGUQ.jl

  • Applies Bayesian Information Gap Decision Theory to uncertainty and decision analysis.
  • Supports experimental-design and robustness questions within MADS workflows.
  • Includes an example comparing source-termination decisions through robustness curves.
Parameter distributions and pairwise sample relationships for AffineInvariantMCMC.jl.
Inference & uncertaintyPublic scientific code

AffineInvariantMCMC.jl

  • Implements the Goodman-Weare affine-invariant ensemble MCMC sampler.
  • Uses multiple walkers to explore Bayesian parameter distributions.
  • Provides a documented Julia sampling example and integration with the MADS ecosystem.