AI INFRASTRUCTURE WITH AUSTRALIAN-CONTROL DEFAULTSAI control defaults

RunForge — AI systems that fit your controls, not vendor defaults

Deploy private LLMs, coding-agent rollout, and workflow automation with data flow, provider exceptions, and ownership documented before rollout.

Book enterprise review call View AI assessment

Australian-region defaults
Provider exceptions documented
Controls and handover

Built for governed rollout, reviewable artifacts, and scoped delivery

Start with the control boundary, then confirm the paid AI readiness assessment before reviewing services, pricing, and rollout detail.

Control boundary first
AU-region self-hosted defaults are separated from hosted exceptions.Self-hosted Australian-region defaults are separated from provider-hosted, SaaS, or broader APAC exceptions before tool decisions.
Artifacts before rollout
Data-flow maps, exception registers, controls, and runbooks come first.Data-flow maps, exception registers, rollout controls, and runbooks are produced before production AI use expands.
Commercial path stays scoped
Assessment or fixed-price delivery comes before operating commitments.Start with assessment or fixed-price delivery, then decide operation and ownership once controls are understood.

AI operating model

Deployment modes before tool selection

Choose the control boundary first; exceptions get written down.Choose the operating model before choosing tools. The default is controlled infrastructure; exceptions are explicit, documented, and tied to the use case.

AI assessment Service paths

DefaultDefault
Self-hostedSelf-hosted by default
AU regions in infrastructure you control.Private LLMs, custom agents, and n8n-based workflow automation are scoped to run in Australian regions inside infrastructure you control unless a documented exception is approved.
ExceptionException path
ExceptionsApproved exception when needed
Hosted or APAC use is documented first.If a provider-hosted tool or broader APAC footprint is the right tradeoff, we document the data flow, controls, and approval path up front.
GovernGovernance
Tool policyTool governance stays explicit
Telemetry, policy, and owners are reviewed.Developer tooling and provider-hosted workflow products can emit telemetry. We assess that risk before rollout and write it down.

View full operating-model detail

Self-hosted by default

Private LLMs, custom agents, and n8n-based workflow automation are scoped to run in Australian regions inside infrastructure you control unless a documented exception is approved.

Approved exception when needed

If a provider-hosted tool or broader APAC footprint is the right tradeoff, we document the data flow, controls, and approval path up front.

Tool governance stays explicit

Developer tooling and provider-hosted workflow products can emit telemetry. We assess that risk before rollout and write it down.

Rollout controls

Governance before production use

Pilot, policy, review, logging, rollback, and owners agreed first.

SSO and RBAC mapped before rollout
Approved-tool policy and exception register
Prompt and data classification guidance

View rollout controls

SSO and RBAC mapped before rollout
Approved-tool policy and exception register
Prompt and data classification guidance
Human approval gates for agentic actions
Audit trail expectations and repository permissions
Disable, rollback, and escalation plan

View 3 more controls

Human approval gates for agentic actions
Audit trail expectations and repository permissions
Disable, rollback, and escalation plan

SSO and RBAC mapped before rollout
Approved-tool policy and exception register
Prompt and data classification guidance
Human approval gates for agentic actions
Audit trail expectations and repository permissions
Disable, rollback, and escalation plan

Start here

AI Readiness Assessment

$10,000AUD

A senior AI engineer assesses your organisation's AI readiness, maps practical opportunities, audits data-flow and sovereignty assumptions, and delivers a prioritised roadmap with clear implementation options.

Acceptance: Each paid assessment SOW names the required readout artifacts, acceptance criteria, and remedy path before work starts, including the ranked risk map, prioritised roadmap, and fixed-scope next options.

Commercial model

Fixed scope

Timeline

1-2 weeks

Best for: Teams exploring AI adoption who need an independent assessment of opportunities, risks, and the most effective path to production AI

Book AI assessment call

Assessment output

Fixed scope

First 30-day outcome

Your first 30 days become a governed AI delivery path

The assessment separates useful AI opportunities from sovereignty, tooling, and ownership risks before any production rollout begins.

Approved use cases ranked by value, data exposure, and operational load
Provider-hosted exceptions documented before adoption decisions
Fixed-price implementation options for the highest-priority initiatives

What you get

AI opportunity audit across engineering and operations

Data sovereignty and compliance gap analysis

Current AI tool usage audit (shadow AI risk assessment)

View remaining outputs+4

Prioritised implementation roadmap with commercial context
Technology stack recommendation (self-hosted vs hosted)
Executive summary for leadership team
Fixed-price proposals for top 3 recommended initiatives

AI opportunity audit across engineering and operations

Data sovereignty and compliance gap analysis

Current AI tool usage audit (shadow AI risk assessment)

Prioritised implementation roadmap with commercial context

Technology stack recommendation (self-hosted vs hosted)

Executive summary for leadership team

Fixed-price proposals for top 3 recommended initiatives

Services menu

AI services built for production

Choose the service path after the controls, data-flow, and ownership model are clear.Every project is fixed-price. Scoped after a 30-minute discovery call, with a proposal target once there is enough discovery context and the operating controls documented before rollout.

Prices are AUD ex GST unless stated and exclude cloud, GPU, SaaS, model-provider, licence, and usage charges paid directly by your organisation.

Decision path
Govern
Controls
Use-case fit, data-flow review, provider exceptions, and coding-agent rollout controls.
View path
Decision path
Deploy
Systems
Workflow automation, private inference, custom AI apps, and production handover.
View path
Decision path
Operate
Agents
Agent guardrails, runbooks, escalation paths, and measurable ownership after rollout.
View path

Govern2-6 weeksFrom $12,000

Govern path: AI Developer Tooling & Coding Agents

Codex · Claude Code · Cursor · OpenCode

Roll out AI coding tools with SSO, policy controls, telemetry review, repo skills, permission profiles, evals, Browser QA, PR handoff, and agent configuration. Provider-hosted data-flow tradeoffs are called out before rollout.

View scope and controls

Tool assessment
SSO & policy setup
Agent permissions
Repo rules & skills

Scope fixed-price proposal

Tool assessmentSSO & policy setupAgent permissions+5 more

Get a fixed-price proposal

DeployQuickest start2-4 weeksFrom $5,000

Deploy path: AI Workflow Automation

n8n Self-Hosted · Integrations · LLM Workflows

Deploy includes workflow automation, private inference, and custom AI applications once the data-flow and operating controls are approved.

Map a manual process, design AI-assisted workflows, and deploy self-hosted n8n by default. Provider-hosted exceptions are documented before rollout. Deliverables include the running automation environment, configured workflows, checkpoints, and team training.

View scope and controls

Process mapping workshop
Workflow architecture & design
AI prompt engineering
Multi-system integration

Scope fixed-price proposal

Process mapping workshopWorkflow architecture & designAI prompt engineering+5 more

Get a fixed-price proposal

Deploy3-6 weeksFrom $8,000

Deploy path: Private LLM Inference Hosting

vLLM · Ollama · TGI · Multi-Cloud

Deploy self-hosted models in your cloud account or an approved provider environment. The scope covers API compatibility, monitoring, load testing, control documentation, and an operations runbook.

View scope and controls

Deploy self-hosted models in your cloud account or an approved provider environment. The scope covers API compatibility, monitoring, load testing, control documentation, and an operations runbook.

GPU infrastructure provisioning
Model deployment & optimisation
OpenAI-compatible API
Load testing & tuning

Scope fixed-price proposal

GPU infrastructure provisioningModel deployment & optimisationOpenAI-compatible API+3 more

Get a fixed-price proposal

Deploy3-16 weeksFrom $20,000

Deploy path: Custom AI Agents & Applications

RAG · Multi-Agent · Production-Grade

Build a purpose-specific AI application, RAG pipeline, or multi-agent workflow on infrastructure you control, with API documentation, deployment scripts, and maintenance runbooks.

View scope and controls

Build a purpose-specific AI application, RAG pipeline, or multi-agent workflow on infrastructure you control, with API documentation, deployment scripts, and maintenance runbooks.

Use case discovery & scoping
RAG pipeline development
Tool & API integrations
Human-in-the-loop controls

Scope fixed-price proposal

Use case discovery & scopingRAG pipeline developmentTool & API integrations+3 more

Get a fixed-price proposal

Operate3-8 weeksFrom $15,000

Operate path: Bounded Agent Workflows

CrewAI · LangGraph · n8n

Design and deploy bounded agents for approved workflows with human review, scoped permissions, rollback paths, guardrail configuration, and operational handover.

View scope and controls

Design and deploy bounded agents for approved workflows with human review, scoped permissions, rollback paths, guardrail configuration, and operational handover.

Workflow discovery & value mapping
Agent architecture design
Multi-platform integration
Custom skill development

Scope fixed-price proposal

Workflow discovery & value mappingAgent architecture designMulti-platform integration+3 more

Get a fixed-price proposal

Need a bundled package? Combine services into one delivery plan. See bundled pricing

AI services pricing

Fixed pricing for staged AI scopes

Book enterprise review call

Larger AI programmes are sequenced after assessment, governance review, and leadership approval. Prices in AUD ex GST unless stated.

Service option	Entry Governed pilot or first workflow	Mid-range Production hardening and integrations	Programme Broader scope after assessment	Retainer Ongoing advisory and support
AI Workflow Automation	Starting from $5,000	$25,000-35,000	$60,000	$2,500-4,000/mo
Private LLM Inference Hosting	Starting from $8,000	$28,000-55,000	$90,000	$4,000-8,000/mo
AI Developer Tooling & Coding Agents	Starting from $12,000	$30,000	$65,000	$3,000-5,000/mo
Bounded Agent Workflows	Starting from $15,000	$35,000-45,000	$150,000	$4,000-6,000/mo
Custom AI Agents & Applications	Starting from $20,000	$40,000-70,000	$120,000	$4,000-6,000/mo

AI Workflow Automation

Starting from $5,000

Mid-range: $25,000-35,000
Programme: $60,000
Retainer: $2,500-4,000/mo

Private LLM Inference Hosting

Starting from $8,000

Mid-range: $28,000-55,000
Programme: $90,000
Retainer: $4,000-8,000/mo

AI Developer Tooling & Coding Agents

Starting from $12,000

Mid-range: $30,000
Programme: $65,000
Retainer: $3,000-5,000/mo

Bounded Agent Workflows

Starting from $15,000

Mid-range: $35,000-45,000
Programme: $150,000
Retainer: $4,000-6,000/mo

Custom AI Agents & Applications

Starting from $20,000

Mid-range: $40,000-70,000
Programme: $120,000
Retainer: $4,000-6,000/mo

Book enterprise review call

Bundled packages

Combine services into one delivery plan

After the operating model is clear, packages reduce sequencing risk and align with the assessment-led roadmap.

Packaged scope

AI-Ready Engineering Team

From $40,000

Developer Tooling + Private LLM Inference

Retainer: $6,000-10,000/mo

Combined delivery

Sovereign AI Stack

$50,000-75,000

Bounded Agents + Workflow Automation

Retainer: $5,500-8,000/mo

Assessment-led programme

Phased AI Adoption Programme

$90,000-180,000

Multiple services sequenced after assessment

Retainer: $10,000-20,000/mo

AI-Ready Engineering Team

Packaged scope

Developer Tooling + Private LLM Inference

Equip your engineering team with AI coding tools and a private LLM endpoint. Private inference runs on Australian infrastructure; provider-hosted tooling paths are documented before rollout.

Package detail

Equip your engineering team with AI coding tools and a private LLM endpoint. Private inference runs on Australian infrastructure; provider-hosted tooling paths are documented before rollout.

From $40,000

Retainer: $6,000-10,000/mo

Sovereign AI Stack

Combined delivery

Bounded Agents + Workflow Automation

AI agents and automated workflows run in your environment by default, with provider API or APAC exceptions documented before rollout.

Package detail

AI agents and automated workflows run in your environment by default, with provider API or APAC exceptions documented before rollout.

$50,000-75,000

Retainer: $5,500-8,000/mo

Phased AI Adoption Programme

Assessment-led programme

Multiple services sequenced after assessment

A controlled adoption path for teams that need developer tools, private inference, agents, applications, and workflow automation sequenced over time. Final scope follows assessment, governance review, and leadership approval.

Package detail

$90,000-180,000

Retainer: $10,000-20,000/mo

Runtime ownership, data handling, and incident paths are written into scope before production rollout.

Runtime ownership

Support ownership, monitoring scope, alert routes, and fallback behaviour are written into the SOW before production rollout.

Data handling

Prompt, output, log, backup, and telemetry handling is mapped before deployment, including provider-hosted exceptions.

Incident path

AI workloads use agreed severity definitions, response targets, rollback options, and named client escalation contacts.

Prices are AUD ex GST unless stated and exclude cloud, GPU, SaaS, model-provider, licence, and usage charges. Every engagement begins with a free 30-minute call. We scope your needs and provide a fixed-price proposal target once there is enough discovery context.

AI developer tooling

Coding agents without unmanaged tool sprawl.

Roll out coding agents with policy, permission, evaluation, and handover controls.Roll out coding agents as an engineering operating model: policies, permissions, repo skills, evals, Browser QA, PR handoff, and a handover path your team can keep.

Provider-hosted tools need approved exceptions

Hosted coding tools can process prompts, repo context, diffs, logs, or telemetry. Approve scope and exceptions first.

Provider-hosted tooling note

Some developer tools and workflow platforms are provider-hosted and may send telemetry or prompt data to their vendors. We help you evaluate those flows and keep exceptions explicit.

Tooling rollout review

Share your coding-agent stack and we will map the first permissions, data-flow, and handover controls to review.

Ask about tooling rollout

Example rollout matrix

Tooling surfaces mapped to controls and handover.

Surface

Codex

Controls configured

Repo-local instructions, sandbox and permission rules, Browser QA, review handoff, and validation commands.

Handover artifact

AGENTS map, repo skills, allowed command profile, eval checks, and PR evidence template.

Surface

Claude Code

Controls configured

Settings, local skills, subagent boundaries, command permissions, and artifact-producing review loops.

Handover artifact

Shared skill parity, reviewer prompts, expected-state notes, and training guide for engineers.

Surface

Cursor

Controls configured

Workspace rules, model/tooling policy, sensitive-path exclusions, and review-ready development workflow.

Handover artifact

Ruleset, onboarding notes, exception register, and team workflow recommendations.

Surface

OpenCode

Controls configured

Project config, explicit allow/deny command rules, agent parity, and command workflows.

Handover artifact

Config review, mirrored agent definitions, command reference, and rollback notes.

AI developer tooling surfaces mapped to configured controls and handover artifacts.
Surface	Controls configured	Handover artifact
Codex	Repo-local instructions, sandbox and permission rules, Browser QA, review handoff, and validation commands.	AGENTS map, repo skills, allowed command profile, eval checks, and PR evidence template.
Claude Code	Settings, local skills, subagent boundaries, command permissions, and artifact-producing review loops.	Shared skill parity, reviewer prompts, expected-state notes, and training guide for engineers.
Cursor	Workspace rules, model/tooling policy, sensitive-path exclusions, and review-ready development workflow.	Ruleset, onboarding notes, exception register, and team workflow recommendations.
OpenCode	Project config, explicit allow/deny command rules, agent parity, and command workflows.	Config review, mirrored agent definitions, command reference, and rollback notes.

Example artifact

Agent permission matrix

Maps tools, commands, sensitive paths, approval gates, and who can widen access.

Example artifact

Repo skills and rules

Keeps project-specific workflow, review, deployment, and section guidance close to the code.

View 4 more artifacts

Evaluation log
Records representative prompts, expected behaviour, drift checks, and reviewer findings.
Browser QA checklist
Defines rendered desktop/mobile checks for agent-made UI changes before PR handoff.
PR handoff pack
Summarises ExecPlan link, validation evidence, residual risk, and rollback notes.
Rollout controls
Defines pilot scope, excluded repositories, provider-hosted exceptions, disable path, and adoption cadence.

Map
Current tooling and data flow
Tool inventory, telemetry notes, and sensitive repository exclusions.
Guard
Rules, permissions, and evals
Repo skills, command policies, approval gates, and drift checks.
Pilot
Browser QA and PR handoff
Review loop, screenshot evidence, PR template, and team feedback.
Handover
Operate without RunForge
Owner map, runbook, rollback path, and update cadence.

Data Sovereignty

Designed for Australian-region self-hosting

For self-hosted AI work scoped in the SOW, Australian cloud regions are the default. Provider-hosted tooling, broader APAC deployment, telemetry, support access, backups, and hosted model APIs are documented as explicit exceptions before rollout.

Control and exception evidence

Australian cloud regions as standard
Self-hosted deployments are scoped for Australian regions (Sydney, Melbourne) by default. New Zealand, Singapore, broader APAC, SaaS tooling, and hosted model APIs are documented as exceptions when your requirements allow.
VPC isolation
For self-hosted deployments, models run inside your VPC or equivalent private network. RunForge does not place those workloads on a shared RunForge inference tenant; provider-hosted exceptions are documented separately.
Evidence-ready documentation
Deployment handover includes architecture notes, data-flow assumptions, control mapping, and runbooks aligned to the requirements you nominate.

Some developer tools and workflow platforms are provider-hosted and may send telemetry or prompt data to their vendors. We help you evaluate those flows and keep exceptions explicit.

Regulatory references are provided for general awareness. Specific obligations depend on your entity type and classification. Consult your legal and compliance team.

Regulatory framework

Australian privacy law imposes conditions on cross-border disclosure of personal information. Regulated industries can face additional requirements under the SOCI Act 2018, APRA CPS 234, APRA CPS 230, and health-record rules.

Privacy Act 1988
APP 8: Imposes conditions on cross-border disclosure of personal information
SOCI Act 2018
Part 2A: Critical Infrastructure Risk Management Program obligations for systems of national significance
APRA CPS 234
Information security standards for regulated financial entities
APRA CPS 230
Operational risk and service provider management obligations for APRA-regulated entities
My Health Records Act 2012
Health record handling obligations for the My Health Record system

Healthcare
Financial Services
Government
Legal
Defence
Education
Transport & Logistics

Self-hosted Australian regions
Approved APAC exceptions documented
VPC isolation
Audit-ready documentation

Technology coverage

Fluent across the AI and infrastructure ecosystem

Current-state review without forced migration. Coverage groups show review fluency, not vendor endorsement, certification, or customer-logo proof.

Private inference
Review vLLM, Ollama, Hugging Face, CUDA, data boundaries, and Australian-region runtime choices.
Developer tooling
Roll out Codex, Claude Code, Cursor, and OpenCode with repo rules, evals, permissions, and PR handoff.
Agent workflows
Scope CrewAI, LangGraph, and n8n workflows behind approvals, rollback paths, and human review.
Platform substrate
Keep Terraform, Docker, and Kubernetes ownership visible so AI work lands inside operable infrastructure.

Next sensible step

Ready to deploy AI on your terms?

Book an enterprise review call. We'll scope what makes sense for your team and give you a fixed-price proposal target once there is enough discovery context.

Book enterprise review call

RunForge — AI systems that fit your controls, not vendor defaults

Built for governed rollout, reviewable artifacts, and scoped delivery

Control boundary first

Artifacts before rollout

Commercial path stays scoped

Deployment modes before tool selection

Self-hostedSelf-hosted by default

ExceptionsApproved exception when needed

Tool policyTool governance stays explicit

Self-hosted by default

Approved exception when needed

Tool governance stays explicit

Governance before production use

AI Readiness Assessment

Your first 30 days become a governed AI delivery path

AI services built for production

Govern

Deploy

Operate

Govern path: AI Developer Tooling & Coding Agents

Deploy path: AI Workflow Automation

Deploy path: Private LLM Inference Hosting

Deploy path: Custom AI Agents & Applications

Operate path: Bounded Agent Workflows

Fixed pricing for staged AI scopes

AI Workflow Automation

Private LLM Inference Hosting

AI Developer Tooling & Coding Agents

Bounded Agent Workflows

Custom AI Agents & Applications

Combine services into one delivery plan

AI-Ready Engineering Team

Sovereign AI Stack

Phased AI Adoption Programme

AI-Ready Engineering Team

Sovereign AI Stack

Phased AI Adoption Programme

Runtime ownership

Data handling

Incident path

Coding agents without unmanaged tool sprawl.

Tooling surfaces mapped to controls and handover.

Codex

Claude Code

Cursor

OpenCode

Codex

Claude Code

Cursor

OpenCode

Agent permission matrix

Repo skills and rules

Evaluation log

Browser QA checklist

PR handoff pack

Rollout controls

Current tooling and data flow

Rules, permissions, and evals

Browser QA and PR handoff

Operate without RunForge

Designed for Australian-region self-hosting

Australian cloud regions as standard

VPC isolation

Evidence-ready documentation

Privacy Act 1988

SOCI Act 2018

APRA CPS 234

APRA CPS 230

My Health Records Act 2012

Fluent across the AI and infrastructure ecosystem

Ready to deploy AI on your terms?