The AIEP Ecosystem

AIEP is designed as a complete AI reasoning infrastructure, not a single software system. The ecosystem consists of multiple interconnected layers that together provide a governed cognitive architecture capable of supporting AI applications, enterprise systems, regulated deployments, autonomous research agents, and future AGI architectures.

Ecosystem Overview

┌─────────────────────────────────────────────────────────────┐
│                      APPLICATIONS                           │
│    Forecast · AIEP Miner · Enterprise Tools · Research      │
└────────────────────────────┬────────────────────────────────┘
                             │
┌────────────────────────────▼────────────────────────────────┐
│                       AGENT LAYER                           │
│              PIEA (Protocol Intelligence Evidence Agent)    │
│     evidence-grounded reasoning · dissent · chain replay    │
└────────────────────────────┬────────────────────────────────┘
                             │
┌────────────────────────────▼────────────────────────────────┐
│                    REASONING LAYER                          │
│  Replayable Chains · Branch Detection · Dissent Signals     │
│          Hypothesis Forking · Multi-Model Reasoning         │
└────────────────────────────┬────────────────────────────────┘
                             │
┌────────────────────────────▼────────────────────────────────┐
│                    EVIDENCE LAYER                           │
│  Discovery · Hashing · Mirror Generation · Trust Scoring    │
│                Evidence Indexing · Retrieval                │
└────────────────────────────┬────────────────────────────────┘
                             │
┌────────────────────────────▼────────────────────────────────┐
│                    KNOWLEDGE LAYER                          │
│  Entity Graphs · Temporal Reassembly · Persistent World     │
│           Models · Jurisdictional Segmentation             │
└────────────────────────────┬────────────────────────────────┘
                             │
┌────────────────────────────▼────────────────────────────────┐
│                    GOVERNANCE LAYER                         │
│  Confidence Tiers · Evidence Sufficiency · Compliance       │
│         Certificates · Meta-Governance Reasoning           │
└────────────────────────────┬────────────────────────────────┘
                             │
┌────────────────────────────▼────────────────────────────────┐
│                 INFRASTRUCTURE LAYER                        │
│  Sovereign Knowledge Nodes · Mirror Networks · Evidence     │
│           Index · Protocol Registries                      │
└────────────────────────────┬────────────────────────────────┘
                             │
┌────────────────────────────▼────────────────────────────────┐
│                    HARDWARE LAYER                           │
│       Trusted Node Hardware · Attestation · Secure          │
│                   Evidence Storage                          │
└─────────────────────────────────────────────────────────────┘

Layer-by-Layer Reference

Applications

The application layer sits at the top of the AIEP stack. Applications interact with PIEA agents through the AIEP protocol. Two running application demonstrations:

• Forecast — a construction governance SaaS built on AIEP evidence-grounded reasoning
• AIEP Miner — an open-source reference implementation showing governed autonomous control of physical hardware. Every decision, every suppressed action, and every failure is committed as tamper-evident evidence to an append-only ledger.

Other application types include enterprise knowledge assistants, regulatory compliance systems, and research platforms.

Agent Layer — PIEA

PIEA (Protocol Intelligence Evidence Agent) is the AIEP reference agent. It implements the full AIEP Piea Surface (P116–P128) in production.

PIEA capabilities:

• evidence-grounded reasoning — every response requires an evidence commitment before generation
• dissent generation — insufficient evidence produces a dissent record, not a fabricated answer
• semantic interpretation branching — ambiguous queries produce both interpretations with shared evidence
• reasoning chain replay — any reasoning step can be replayed deterministically
• autonomous research loops — the agent can initiate evidence-gathering cycles to resolve knowledge gaps

PIEA proves the AIEP architecture: it cannot hallucinate in the structural sense because the protocol does not permit generation without evidence.

Reasoning Layer

Governs how AI inference occurs. Key mechanisms:

Evidence Layer

Manages the data sources used in reasoning. The evidence layer creates a persistent knowledge substrate that outlives individual queries.

Knowledge Layer

Organises evidence into structured knowledge with temporal depth.

Governance Layer

Ensures reasoning reliability and enables regulated deployment.

AIEP compliance certificates include response hash, reasoning chain hash, evidence hashes, model identifiers, and jurisdiction scope — allowing independent verification without access to model weights.

Infrastructure Layer

Supports distributed deployment at network scale.

Hardware Layer

The AIEP hardware layer (P91, P99, P100) defines the trust substrate for governance-enforced AI nodes. The aiep-node reference emulator is published as open-source developer tooling; actual hardware implementation details are proprietary.

The hardware layer enables:

• cryptographic attestation of node identity
• governance-enforced execution environments
• secure evidence storage with tamper evidence
• cross-node consensus without a trusted coordinator

Open-Source Repositories

The following repositories implement portions of the AIEP ecosystem stack. Repositories marked Planned have not yet been published — they are roadmap items and will be released as the protocol matures.

All published repositories are released under the Apache License 2.0.

Multi-Model Reasoning

AIEP systems are model-agnostic. Multiple model types can be coordinated:

• large language models
• structured reasoning engines
• symbolic reasoning systems
• domain-specific fine-tuned models

Model diversity increases reasoning robustness and prevents single-point model bias.

Autonomous Research

AIEP agents may initiate autonomous research cycles to resolve knowledge gaps:

Hypothesis Generation
        ↓
Evidence Search
        ↓
Analysis
        ↓
Branch Evaluation
        ↓
Conclusion
        ↓
World Model Update

This allows AIEP systems to continuously expand their knowledge base without relying on model training cycles.

Scale and Strategic Context

The AIEP ecosystem introduces a new infrastructure category: evidence-governed reasoning networks.

This infrastructure differs from existing AI systems by combining:

• knowledge preservation (content-hashed evidence substrate)
• reasoning transparency (replayable chains, dissent records)
• governance mechanisms (formal policy evaluation, certified outputs)
• distributed infrastructure (mirror networks, federated knowledge exchange)

At global scale, this infrastructure supports: distributed research agents, cross-node knowledge synthesis, shared reasoning infrastructure, and multi-jurisdictional compliance verification.

Further Reading

• Architecture — full technical layer reference
• Specification Architecture — how specs map to layers
• Open Specifications — browse all 171 open-source specifications
• Developer Guide — build with AIEP
• PIEA — the reference agent implementation
• AIEP Miner — governed autonomous hardware control, open source