How Does Olivares AI Differ From An LLM Gateway?

A gateway governs model calls such as routing, keys and caching, while observability tools handle traces and prompts. Olivares AI sits around the estate instead, mapping which agents touch which resources and comparing permitted grants with observed behaviour. Both layers can run alongside it.

Can Olivares AI Break Production Traffic?

The discovery path is out of band. It reads audit logs, OpenTelemetry and source-native signals rather than sitting inside every agent data path, so a collector failure degrades visibility only. Enforcement applies solely where you deliberately wire a governed gate, and those gates fail closed.

What Does Self-Hosted Mean If Claude Runs In The Cloud?

Claude inference still reaches Anthropic or your configured cloud account. What stays inside your perimeter is Olivares itself: discovery, the access map, audit ledger, policy state and inventory. That governance layer can operate fully air-gapped with no telemetry sent back.

How Reliable Is The Read Versus Write Distinction?

Every edge carries two honesty axes: how well the source proves read versus write, and how firmly the access ties to one agent. Where a shared service account hides identity, the product marks the uncertainty rather than guessing, prompting better instrumentation.

Which Data Sources Can Olivares AI Ingest Today?

Supported connectors include PostgreSQL via pgAudit, AWS CloudTrail, S3, Kubernetes, Vault, Kafka, MCP servers and an eBPF observer for file and network I/O. The Apache-2.0 connector SDK lets teams build additional sources without copyleft friction.

Listed · Reviewed Aug 2026 · 7-criteria rubric

AI Productivity Tools

Olivares AI

Name: Olivares AI
Availability: InStock
Author: Olivares AI (Francisco Olivares)

Self-Hosted Governance Control Plane For Enterprise AI Agent Estates

by Olivares AI (Francisco Olivares) · Consuegra, Toledo, Spain · Founded 2025

Visit Olivares AI →

Pricing

Contact Sales

What is Olivares AI?

Olivares AI is a self-hosted, vendor-neutral control plane that discovers, governs and audits the AI agents running on enterprise infrastructure, built around Claude and Claude Code.

It ships as a single static Go binary with an embedded console, passively inventorying agents, sessions, models and MCP servers, then building a read/write access map that diffs permitted against observed behaviour.

Platform, DevOps and security teams use it to enforce deny-closed policy, attribute model spend and export tamper-evident audit evidence without governance data ever leaving their own perimeter. Open-core under AGPL-3.0, currently pre-1.0.

Olivares AI — Self-Hosted Governance Control Plane For Enterprise AI Agent Estates Whether you're evaluating Olivares AI for your team or comparing it to alternatives in the AI Productivity Tools category, this in-depth review covers everything: features, pricing, real user reviews, pros and cons, integrations, and direct comparisons against competitors.

Key Features 10

Self-hosted, vendor-neutral AI governance control plane built around Claude and Claude Code

Passive agent discovery inventories every AI agent, session, model and MCP server

Read/write access map comparing permitted grants against actually observed agent behavior

Policy-as-code guardrails enforced at access time, deny-closed with least-privilege paths

Append-only, hash-chained audit ledger exportable as PDF, CSV or JSON evidence

Real-time AI cost attribution by agent, model and team with budget enforcement

Human-in-the-loop approvals, risk tiers, dual control and audited break-glass access

mTLS between components, read-only collector, air-gapped operation with zero calls home

Who Is Olivares AI For

1 Platform Engineering Teams

2 DevOps And SRE Engineers

3 Enterprise Security Architects

4 CISOs And Compliance Officers

5 FinOps And Cloud Cost Analysts

6 Self-Hosting Systems Administrators

Pros & Cons

Pros 4 benefits

Air-Gapped, Zero Calls Home
Coverage And Attribution Confidence Labelling
Single Static Go Binary
Signed SLSA Provenance And SBOM

Cons 3 limitations

Pre-1.0, No Tagged Release
AGPL Copyleft Blocks Some Policies
Commercial Pricing Not Yet Published

Frequently Asked Questions

5 questions

Who is Olivares AI for?

Olivares AI is most useful for Platform Engineering Teams, DevOps And SRE Engineers, Enterprise Security Architects and CISOs And Compliance Officers.

Olivares AI pricing

Olivares AI uses contact-based (enterprise) pricing. Contact Sales. For the current tier breakdown and any limits, see the pricing section above or check the vendor's pricing page directly — limits and prices change.

User Base

Pre-release, early access

Active Users

Security & Privacy

Customer-controlled; self-hosted on-prem, multi-cloud or air-gapped

GDPR Aligned (Self-Hosted Controller Model) ISO/IEC 27001:2022 Readiness-Mapped (Not Certified) ISO/IEC 42001:2023 Readiness-Mapped (Not Certified) SOC 2 Type II Readiness-Mapped (Not Attested) EU AI Act Annex IV Documentation Mapping NIST AI RMF 1.0 Subcategory Mapping

mTLS on every internal control-plane hop Append-only hash-chained tamper-evident audit ledger Cedar policy engine with role and attribute-based access control WebAuthn, PIV/CAC and AAL step-up authentication BYOK encryption, DLP and gVisor/Firecracker sandbox isolation Cosign keyless signatures, SLSA Build Level 3 provenance and SBOM +2 more

Collaboration & Teams

Team Workspaces Multi-User Access Role Permissions Shared Projects Version History Activity Log

Learning & Support

Resources

Documentation Templates

Community

Forum

Support Channels

Email Priority Dedicated Manager Onboarding

Localization

UI Languages

Content Languages

Multi-Language UI

Recognition & Trust

Open Source

All Features of Olivares AI