What Xtructure does

Xtructure LLC is a privately held artificial intelligence research company founded in 2001 and domiciled in the United States, operating with a sole founder-operator structure (Michael Roberts, Founder & Lead Researcher). The company develops algorithms, tools, and theory at the intersection of evolutionary computation and deep learning, with a stated focus on applying genetic algorithms to neural network topology, weights, and learning rules as an alternative to gradient descent, plus discrete event simulation environments and novel hypergraph representations for structured knowledge.

Its product portfolio consists of two core offerings. XNet is a Python platform for discrete event simulation and computational evolution, providing a composable toolkit for population management, genotype/phenotype encoding, fitness evaluation, and simulation scheduling; it was originally released in Java as open source in 2012 and subsequently ported to Python for Jupyter Notebook use. XBundle is a hierarchical hypergraph database patented in 2025, designed to represent and query complex, multi-relational data structures with edges spanning arbitrary sets of nodes, positioning it as a more expressive alternative to conventional binary graph databases.

Xtructure's business model is built on open-source distribution rather than commercial licensing. XNet is released under the MIT License on GitHub with no disclosed pricing, no enterprise sales motion, and no disclosed paying customers or revenue; the company identifies a single horizontal target segment — the AI Research Community. No funding rounds, external investors, partnerships, acquisitions, or commercial channels are disclosed, and ownership appears concentrated with the founder.

Name

Xtructure

Legal name

Xtructure LLC

Website

https://xtructure.com

Company type

Private

Founded year

2001

Operating status

Operating

Headcount range

1–10 employees

Short description

Xtructure LLC is a privately held U.S. AI research company founded in 2001 that develops evolutionary-computation tools (XNet) and a patented hypergraph database (XBundle), distributing XNet as open source to the AI research community.

Ownership category

akta.pro rank

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Operations, Marketing or Sales

Distribution channels1 record

Marketing channels1 record

Core offering

Xtructure is an artificial intelligence research company that develops algorithms, tools, and theory at the intersection of evolutionary computation and deep learning. Its two core offerings are XNet, a Python platform for discrete event simulation and computational evolution (released as open source in 2012), and XBundle, a patented hierarchical hypergraph database (2025) designed to represent and query complex multi-relational data structures beyond the expressiveness of conventional graph databases.

Product overview

Xtructure is an artificial intelligence research company developing algorithms, tools, and theory at the intersection of evolutionary computation and deep learning. The company offers two core products: XNet (a Python platform for discrete event simulation and computational evolution) and XBundle (a hierarchical hypergraph database for representing complex, multi-relational data structures). XNet was originally released as open source in Java in 2012 and subsequently ported to Python, while XBundle received patent protection in 2025. Together, these products enable evolutionary strategies for optimizing neural network topology and represent structured knowledge through novel hypergraph representations.

Differentiator

Problem solved

Functional benefit

Brands

• XNet: A Python platform for discrete event simulation and computational evolution, originally released in Java as open source in 2012, now ported to Python and adapted for use in Jupyter Notebooks. Provides a composable toolkit for constructing evolutionary experiments including population management, genotype/phenotype encoding, fitness evaluation, and simulation scheduling.
• XBundle

Products and services

• XNet XNet is a Python platform for discrete event simulation and computational evolution, providing a composable toolkit for constructing evolutionary experiments including population management, genotype/phenotype encoding, fitness evaluation, and simulation scheduling. Originally released in Java as open source in 2012 and now ported to Python for Jupyter Notebook use, it is targeted at AI researchers and machine learning practitioners working with evolutionary strategies and event-driven systems. Distributed under the MIT License.
• XBundle XBundle is a hierarchical hypergraph database, patented in 2025, designed to represent and query complex, multi-relational data structures that exceed the expressiveness of conventional graph databases. Its hypergraph formalism allows edges to span arbitrary sets of nodes rather than only binary connections, enabling richer representations of structured knowledge for advanced AI systems. Targeted at developers and researchers working with complex relational data.

Segments1 record

Ideal customer profiles1 record

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

AI capability4 records

Feature2 records

Scale indicators1 record

Recent moves4 records

Expansion highlights3 records

Broad incumbents

• ArangoDB: A multi-model database supporting graph, document, and key/value workloads in one engine. ArangoDB is comparable to XBundle as an established multi-model database vendor offering graph capabilities as part of a broader data management platform.
• Google DeepMind: A global AI research lab publishing work that spans gradient-based and evolutionary methods at scale. DeepMind is comparable to Xtructure as a well-funded AI research entity that occasionally validates or applies neuroevolution approaches that XNet seeks to serve.
• TigerGraph: An enterprise graph database platform focused on scalable analytics over highly connected data. TigerGraph is comparable to XBundle as an adjacent data-representation incumbent whose commercial trajectory and developer ecosystem set the bar for any newer entrant in graph and hypergraph data management.
• TensorFlow: Google's flagship open-source deep learning framework offering gradient-based training at industrial scale. It is broadly comparable to Xtructure as one of the incumbent AI development platforms competing for the same research and engineering audience as XNet.
• Neo4j: The leading enterprise property-graph database with deep adoption in knowledge graph and AI-driven application stacks. Neo4j is broadly comparable to XBundle because it defines the graph database category that XBundle's hypergraph formalism is positioned as a more expressive alternative to.
• PyTorch: The dominant open-source deep learning framework maintained by Meta with gradient-based optimization as its core paradigm. PyTorch is comparable to Xtructure because it represents the default alternative against which any non-gradient optimization framework competes for researcher attention.
• OpenAI: A leading AI research and deployment organization operating at the frontier of gradient-based model development. OpenAI is comparable to Xtructure as the dominant AI research brand whose tooling and methodological choices shape the ecosystem in which XNet and XBundle compete for relevance.

Direct peers

• DEAP (Distributed Evolutionary Algorithms in Python): An open-source Python framework for evolutionary computation offering population management, genetic operators, and fitness evaluation. DEAP is the closest direct comparable to XNet because both target the same Python-based evolutionary algorithm research community with overlapping toolchain scope.
• Nevergrad (Meta / Facebook Research): Meta's open-source gradient-free optimization library that includes evolutionary and population-based methods. Nevergrad overlaps directly with XNet on alternative-to-gradient optimization tooling for AI and ML research workflows.
• Memgraph: An open-source in-memory graph database targeting developers with CSDN-style query semantics. Memgraph is the closest open-source, developer-first analog to XBundle because both pursue graph-structured data with a similar open distribution posture, even though Memgraph uses binary edges rather than hyperedges.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat3 records

Key risks6 records

Key highlights6 records

Customer concentration

• GitHubgithub.com/xtructure/xnet-python

Revenue estimate

Valuation estimate

Number of profiles

Profiles1 record

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Xtructure