Our Story

John Harby

Founder & CEO, Autonomic AI, LLC

Overview

John Harby is the Founder and CEO of Autonomic AI, LLC, a research-driven technology company advancing next-generation AI architectures with mathematically grounded accuracy, equivalence, and energy efficiency. His work introduces the Accuracy AI framework, demonstrating significant reductions in computational cost and energy usage while maintaining verified correctness.

With university education focused on pure mathematics, John Harby entered the software industry circa 1990. He has held senior engineering and architectural roles at IBM, Oracle, HP, Intuit, and Lion Bioscience, contributing to distributed systems, service-oriented architecture, and applied artificial intelligence. He is a co-author of the SOA Blueprints specification and has participated in multiple industry standards bodies, including the OASIS SOA Reference Model Technical Committee and several Java JSR expert groups.

His current research focuses on deterministic AI code generation, equivalence-class transformations, and energy-optimized computation, positioning his work at the intersection of mathematics, computer science, and scalable AI systems.

Professional Background

Harby’s career spans enterprise architecture, distributed systems, and applied AI across leading technology organizations:

• HP (e-speak platform): Autonomous service composition and distributed systems

• Oracle: Manufacturing and enterprise data models

• IBM Global Services: Service-oriented architecture and enterprise integration

• Intuit: Compiler and code generation systems for financial applications

• Lion Bioscience: Neural modeling and bioinformatics integration

These experiences inform his current work in self-evolving AI systems and morphic model architectures.

Standards & Industry Contributions

Harby has contributed to multiple standards bodies and industry specifications, where designs are reviewed through multi-organization consensus:

• OASIS SOA-RM (Reference Model): Technical Committee Member

• SOA Blueprints Specification: Co-author

• WS-BusinessActivity Specification: Standards Committee Member

• OMG Workflow Management Facility: Contributor

• JSR-343 (JMS 2.0): Expert Group Participant

• JSR-311 (JAX-RS): Expert Group Participant

• JSR-181 (JAX-WS): Contributor

• JSR-208 (JBI): Contributor

He also presented at the OASIS Symposium (2007, San Diego) on service composition and SOA blueprint patterns.

Research & Publications

Harby’s research focuses on the formal relationship between energy, accuracy, and equivalence in AI systems. His work is available through platforms including Google Scholar and Zenodo.

Key areas include:

• Energy-optimized AI computation

• Deterministic and verifiable code generation

• Equivalence-class transformations and canonical representations

• Morphic and functor-based model architectures

Selected works include:

• Energy, Accuracy, and Equivalence: A Mathematical Framework for AI Optimization

• Two-Level Morphic Reduction: Global Quotient Topology and Local Fiber-Wise Transform

• Homology and Galois Framework for Neural Model Spaces

• Ontology Alignment and Equivalence-Class Transformations

Intellectual Property

Harby is the inventor of multiple U.S. patent filings in applied artificial intelligence, including work on:

• energy-efficient AI computation

• accuracy-constrained code generation

• adaptive event detection

• multi-criteria model selection

These innovations form the foundation of Autonomic AI’s morphic and Six Sigma frameworks.

Current Focus

His current work explores:

• functor-style AI architectures

• equivalence-driven computation

• reduction of redundant inference

• edge and energy-efficient AI deployment

This research aims to redefine how AI systems are built, evaluated, and optimized at scale.

       Links

​

Google Scholar 

GitHub

LinkedIn

ORCID

Zenodo