PIRADON develops pattern-native infrastructure — a class of technology in which information carries its own identity, computes its own relationships, and records its own transformations. The result is a substrate that makes the conventional software stack architecturally redundant.
PIRADON is the originator and custodian of the Zero-Point Infrastructure — a proprietary mathematical substrate for content-addressed, pattern-native computation. All research, all intellectual property, and all licensing rights reside with PIRADON.
The organisation operates as both a research programme and a technology licensor. The research produces validated implementations, verified results, and technical reports. The licensing programme makes those implementations available to partners under terms negotiated per engagement.
PIRADON does not publish implementation details. Intellectual property is protected as trade secret. What is published are results: validated performance figures, verified mathematical properties, and descriptions of what the technology does — not how it does it.
"The compensation stack is not a design choice. It is a consequence of a missing mathematical layer. Provide the missing layer, and the stack becomes unnecessary — not optimised, unnecessary."
The research programme spans three generations of implementation, each applying the same foundational concepts — self-identifying, self-relating, self-transforming information — through progressively more complete mathematical frameworks. The current implementation, built on a finite field substrate, is the first to unify all three properties within a single algebraic structure. Results are validated through working implementations and reproducible benchmarks. The programme is ongoing.
Describes PNPL — the programming layer built on Zero-Point Infrastructure. Applications composed as patterns: no compiler, no build system, no runtime dependencies. 23 KB binary. Hardware targets: x86-64, ARM, embedded, FPGA. Covers the three emergent properties, composition model, and validated performance characteristics.
A PostgreSQL-compatible server implemented as a 1,066-byte composed kernel
on the PNPL substrate. Benchmarked with pgbench
against a live PostgreSQL client connection.
554,404 TPS at 16 clients. Zero failures across all benchmark runs.
No B-trees, no indexes, no query planner. Every predicate O(1).
Describes the Membrane application layer built on ZPI — a relational intelligence substrate for AI agents and enterprise knowledge systems. Full Wikipedia corpus (164 million patterns) ingested in 10.5 minutes, resident in 17 GB. Relationship queries: ~1 nanosecond. Context window cost stays constant regardless of corpus size. In daily operational use. Available to license.
The foundational description of Zero-Point Infrastructure. Establishes the three properties — self-identity, self-relation, self-transformation — and their mathematical basis. Demonstrates that these properties are sufficient to make the conventional software infrastructure stack architecturally redundant. All subsequent PIRADON work builds on this foundation.
Every technology in the PIRADON portfolio is built on the Zero-Point Infrastructure. They differ in abstraction level and application domain, not in mathematical foundation. Shared vocabulary deduplicates automatically. Cross-application queries work immediately.
The foundational substrate. A content-addressed field in which information self-identifies, self-relates, and self-transforms. Every relationship between any two patterns is computable in constant time from their coordinates. No index. No schema. No reconciliation. The seed is 22 public operations. Everything else composes from them.
The programming layer. Applications are composed as pattern graphs —
entities created, related, and combined — without a compiler, linker,
or build system. The runtime binary is 23 kilobytes.
Applications are measured in hundreds of bytes.
Portable .pnpl format
runs unchanged on x86-64, ARM, embedded microcontrollers, and FPGA targets.
The knowledge and intelligence layer. AI agents connect via MCP and gain the ability to absorb corpora, query relational structure, and discover connections that were never explicitly defined. Context window consumption stays at 1–2% regardless of corpus size. The library grows in depth; the cost stays flat. Licensed. In operational use.
A PostgreSQL wire-compatible server composed entirely on the PNPL substrate. Demonstrates that a database — typically hundreds of megabytes of compiled code — can be expressed as a 1,066-byte composed kernel with no B-trees, no indexes, no query planner, and no external dependencies. Validated with the industry-standard pgbench tool.
PIRADON works with partners whose applications are well-matched to pattern-native infrastructure. Engagement structures vary — from equity partnerships where PIRADON holds a direct interest in the partner's success, to technology licences for organisations deploying on their own infrastructure. There are no standard pricing tiers. Each engagement is structured to the partner's deployment model, data environment, and support requirements. Enquiries are welcome from any sector.
Current activity spans enterprise knowledge management, defence and intelligence applications, and infrastructure research. On-premises and air-gapped deployments are available for regulated environments.
Enterprise knowledge management and AI agent memory. Absorb any corpus. Query relational structure that was never explicitly defined. Deployable as a managed cloud instance or a single binary on your own infrastructure. Air-gap compatible. No data leaves your environment unless you choose it to.
membrane.informationpatterns.com ↗PIRADON holds a direct equity interest in Systome. The Zero-Point Infrastructure is the technological foundation of Systome's AI Confidant product for enterprise knowledge management and decision support. Systome packages and brings the technology to market; PIRADON and Systome share in the outcome. Their deployment demonstrates the substrate at organisational scale, with the relational field holding the equivalent of any library and delivering relevant structure to AI reasoning in nanoseconds.
systome.ai ↗The substrate and programming layer for organisations building pattern-native applications. Suitable for data infrastructure replacement, edge and embedded deployments, distributed systems without consensus overhead, and any environment where conventional infrastructure complexity is a constraint.
PIRADON accepts enquiries regarding licensing, research collaboration, technical partnerships, and investment. There is no automated process. Enquiries are read and responded to directly.
For licensing discussions, technical enquiries, partnership proposals, or any other matter — contact directly. Please include a brief description of your organisation and the nature of your interest.