Federated Enterprise Intelligence: an operating model for connecting sovereign domain intelligence into a governed, active enterprise
- 1Independent researcher
Abstract
Organisations are moving from AI assistants that answer questions to AI systems that do work and take action. The hard problem is no longer getting a capable model. It is giving those systems enough trusted context and authority to act sensibly across an organisation whose departments legitimately disagree about what words mean, who is allowed to decide and what counts as proof. The two common answers, a separate set of agents in every function or one central AI platform, either fragment the organisation or flatten the differences that matter.
This paper proposes Federated Enterprise Intelligence (FEI). Each major domain runs its own governed Domain Intelligence System. Those systems work together through a shared architecture built on explicit contracts covering meaning, data, authority, policy and evidence. A Governing Centre enforces enterprise-wide rules and mediates between domains but never decides for a domain inside its own area, and it is audited by functions it cannot control. The paper places FEI against domain-driven design, data mesh, organisational digital twins and federated systems; explains why it is feasible now when earlier attempts at enterprise-wide coherence were not; works through one exchange between two domains in detail; describes how decision authority can move from people to systems over time, one decision type at a time, as the evidence justifies it, while accountability stays with named people; and treats assurance as an exchange with properties that automation must preserve. Its value claims are stated as testable hypotheses, each paired with a measure and the result that would disprove it, and the author says which he expects to hold.
Related
- Enterprise Retrieval-Augmented Generation (RAG) over Federated Cloud Data Lakes
- A novel cloud-enabled enterprise information systems framework for accelerating digital transformation strategies
- Editorial Note: Low-carbon governance, fiscal decentralization, and enterprise green development: Evidence from China
- Graph-based ransomware propagation modelling and early interception in enterprise networks using dynamic attack surface analysis
- Adaptive reconstruction mechanism for enterprise collaborative networks based on multi-agent reinforcement learning
BibTeX
@misc{rackham2026federated,
title = {{Federated Enterprise Intelligence: an operating model for connecting sovereign domain intelligence into a governed, active enterprise}},
author = {William Rackham},
howpublished = {Zenodo},
year = {2026},
month = oct,
doi = {10.5281/zenodo.23388857},
url = {https://zenodo.org/records/23388857},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Federated Enterprise Intelligence: an operating model for connecting sovereign domain intelligence into a governed, active enterprise
AU - William Rackham
PY - 2026
DA - 2026/10/21
PB - Zenodo
DO - 10.5281/zenodo.23388857
UR - https://zenodo.org/records/23388857
AB - Organisations are moving from AI assistants that answer questions to AI systems that do work and take action. The hard problem is no longer getting a capable model. It is giving those systems enough trusted context and authority to act sensibly across an organisation whose departments legitimately disagree about what words mean, who is allowed to decide and what counts as proof. The two common answers, a separate set of agents in every function or one central AI platform, either fragment the organisation or flatten the differences that matter. This paper proposes Federated Enterprise Intelligence (FEI). Each major domain runs its own governed Domain Intelligence System. Those systems work together through a shared architecture built on explicit contracts covering meaning, data, authority, policy and evidence. A Governing Centre enforces enterprise-wide rules and mediates between domains but never decides for a domain inside its own area, and it is audited by functions it cannot control. The paper places FEI against domain-driven design, data mesh, organisational digital twins and federated systems; explains why it is feasible now when earlier attempts at enterprise-wide coherence were not; works through one exchange between two domains in detail; describes how decision authority can move from people to systems over time, one decision type at a time, as the evidence justifies it, while accountability stays with named people; and treats assurance as an exchange with properties that automation must preserve. Its value claims are stated as testable hypotheses, each paired with a measure and the result that would disprove it, and the author says which he expects to hold.
ER -