Kraken Networks builds AI governance, service mesh, and orchestration tools.
Open source, built for depth.
PRODUCTS
Python-first expert system runtime built on CLIPS. Deterministic reasoning with persistent working memory, compliance rule packs, and sub-millisecond latency.
engine = fathom.load("rules/governance.yaml")
result = engine.evaluate({
"agent": "analyst",
"action": "access",
"data_classification": "SECRET"
})
# => Decision(action="deny", reason="clearance insufficient")
Policy-first data broker for multi-agent systems. Three personas govern what agents know — the Librarian routes, the Curator learns, the Gatekeeper enforces.
response = nautilus.query(
intent="threat_analysis",
clearance="SECRET",
session="op-47"
)
# Librarian routes → Gatekeeper enforces → results attested
Full-stack platform for building and governing multi-agent systems. Go backend, React dashboard, DSPy reasoning modules, and CLIPS-based policy governors.
workflow.stage("analyze", {
module: "chain_of_thought",
agents: ["analyst"],
governor: "nist-800-53"
})
workflow.stage("review", {
require_approval: true,
on_failure: "escalate"
})
Agent action governance mesh. Controls what agents do — every action passes through Fathom-powered policy before execution.
bosun.authorize({
agent: "classifier",
action: "escalate",
context: ["session-47", "agent-a:output"]
})
# => Authorization(allowed=true, attestation="eyJ...")
20+ tools built / Herndon, VA
ARCHITECTURE
Kraken Networks builds a layered stack for AI agent governance. Each layer handles a distinct concern — reasoning, data governance, action governance, orchestration — so teams can adopt what they need without buying the whole platform.
At the base, Fathom provides deterministic reasoning using CLIPS, a battle-tested expert system runtime. Define governance policies in YAML, and Fathom enforces them with sub-millisecond latency and full audit trails.
Nautilus is the policy-first data broker. It mediates every data request through three personas — the Librarian routes, the Curator learns, the Gatekeeper enforces — tracking cumulative exposure across sessions so no agent quietly builds a dossier.
Bosun governs what agents do. Every action passes through policy before execution, with context declarations that create a complete audit trail of who did what and why.
Railyard orchestrates the fleet. A full-stack platform with workflow DAGs, DSPy reasoning modules, human approval gates, and real-time monitoring — giving you control over complex agent collaborations.
BLOG
How separating agent work from rules-based routing, plus default-deny gates and a pinned zero-outbound runtime, makes an agent stack auditable behind a one-way diode — grounded in a reference workload that runs fully offline.
Why a single configurable budget cap is a single point of failure for self-modifying agents, and how auto-r-graph proves a second, untouchable stop layer with a test that disables the first.
Direct database access for AI agents is a liability. A data broker with three personas — Librarian, Curator, Gatekeeper — gives you deterministic control over what agents know.