Neural Inverse is Open Source →

Neural Inverse Enterprise

The open-source IDE extended with the governance infrastructure regulated organisations need to say yes to AI development. Developers use the same IDE they can run locally. The Enterprise layer adds compliance enforcement, policy controls, and a full audit trail - running on your infrastructure or hosted by us.

Talk to Sales Read the Docs

The Funnel

Neural Inverse has three tiers. Each builds on the one before it.

OSS (Local IDE) - the full open-source IDE running on your machine. Power Mode, firmware tools, legacy modernisation, 20 LLM providers. Free under Apache 2.0. Download

Cloud - the same IDE in a managed workspace. Private git storage, free OSS models, hardware simulation, pay-as-you-go infrastructure. No local install. Get started

Enterprise - the IDE with the governance layer. Everything in OSS plus Checks Manager, IAM Policy Engine, Enclave, AI Governance controls, and Audit Trail. Runs on your infrastructure (on-premises) or as a dedicated hosted instance at platform.neuralinverse.com.

What Enterprise Adds

The open-source IDE has no governance infrastructure - no compliance enforcement, no org-level controls, no audit trail. That is intentional. Individual developers and teams that do not operate in regulated environments do not need it.

Enterprise adds the governance layer for organisations that do.

FeatureOSSCloudEnterprise
AI coding, Power Mode, agentsYesYesYes
Firmware developmentYesYesYes
Legacy modernisationYesYesYes
20 LLM providers / BYOLLMYesYesYes
Free OSS modelsNoYes-
Managed workspacesNoYesOptional
Checks Manager (GRC)NoNoYes
IAM Policy EngineNoNoYes
EnclaveNoNoYes
AI Governance controlsNoNoYes
Audit TrailNoNoYes
On-premises deploymentYesNoYes
Dedicated hosted instanceNoNoYes

Checks Manager

Real-time GRC compliance enforcement as code is written. Rules run against the AST and language server on every save, plus a background AI reasoning layer for semantic violations. Violations appear inline before code is committed.

Supported frameworks: MISRA C:2012/2023, CERT C, IEC 62304, ISO 26262, IEC 61508, DO-178C, SOC 2, AUTOSAR, and custom frameworks your organisation defines.

Every violation and every suppression is written to the audit trail. Learn more

IAM Policy Engine

A custom ni: action namespace with four-tier scope inheritance: org, project, subproject, user. Each level overrides the one above it. Policies are Allow/Deny statements - the same model as AWS IAM applied to AI development capabilities.

A compliance officer can configure which models specific teams use, which tools Power Mode agents can execute, which framework rules block commits, and which developers have elevated access - all from one policy console. Changes propagate to all connected IDEs within 30 seconds. Learn more

Enclave

A context firewall between the IDE and any LLM. Every piece of context an agent or chat session would send - source files, terminal output, error messages - passes through the Enclave first. High-entropy strings, credentials, PII, and custom patterns are blocked or redacted before the request is made.

Works with external models and with self-hosted BYOLLM endpoints. Even for models running inside your own infrastructure, the scan record exists. Learn more

AI Governance

Org-level control over which models developers can use. Enforcement mode prevents developers from using unapproved models or routing around policy with personal API keys. Agent execution controls set limits on iterations, concurrent sub-agents, and tool-type approvals that apply regardless of developer settings.

When a developer needs access to something policy does not allow, they submit a request from inside the IDE. The approval or denial is logged. Learn more

Audit Trail

An immutable structured log of every AI action - chat messages, agent tool calls, model requests, policy changes, approvals, violations, suppressions. Records include which model was used, what policy was active, and what the Enclave found.

Exportable as a tamper-evident bundle with a chain integrity hash. Suitable for regulatory evidence, internal compliance review, and incident investigation. Learn more

Deployment Options

On-premises - deploy inside your own infrastructure via Helm chart or Docker Compose. All services - console, IAM engine, Checks, Enclave, audit storage - run in your environment. Nothing leaves your network. Air-gapped deployment supported.

Dedicated hosted - a dedicated isolated tenant at platform.neuralinverse.com. Managed by Neural Inverse, isolated from other customers, in the region you specify.

Both options run the same platform. The difference is where the infrastructure lives.

Support

Enterprise includes a dedicated support SLA:

  • Severity 1 (production down): 1 hour acknowledgement, 24/7 coverage
  • Severity 2 (major functionality degraded): 24 business hour acknowledgement
  • Severity 3 (minor issue): 48 business hour acknowledgement

Dedicated Slack channel, onboarding assistance, and architectural guidance included.

Get Started

Enterprise starts with a 90-day paid pilot. Contact sales@neuralinverse.com or use the form below.


Was this page helpful?