Back the rails that make AI accountable.
Arrochar Labs is raising a pre-seed round on a SAFE to take ASP™, the Arrochar Superintelligence Platform, from working software to its first paying customers.
Six products run as one platform, AgentGUARD™ on every agent, twenty-one government services built as modules, and seven government panels a buyer can procure through. The round funds the hardening, certification and distribution of what already runs.
Mark Macfarlane, founder and CEO, MSc BIT. Twenty years advising CXOs on technology strategy across government, aviation, logistics and energy, and an IoT Big Data Analytics Global Award Winner, London. The products are built by an AI engineering workforce with Mark in the loop on every decision.
We are looking for one more founder alongside the capital: someone who opens doors to funding, clients and boards that need AI governance advice today.
- Ask
- US$750k on a SAFE
- Valuation cap
- US$3.0m
- Post-money
- US$3.75m, about 20%
- Minimum ticket
- US$25k
Indicative terms. The SAFE instrument governs.
- Status
- Open for expressions of interest
- Target close
- First half of 2027
- Eligibility
- Sophisticated and professional investors (Corporations Act s708)
© 2025 NVIDIA, the NVIDIA logo, and NVIDIA Inception are trademarks and/or registered trademarks of NVIDIA Corporation in the U.S. and other countries. All other trademarks are the property of their respective owners.
What is live today
Nothing on this page needs to be taken on trust. Click through and use it.
MeshGov services
Twenty-one government services, each demonstrating interactively with the officer's decision held at the gate.
Run a serviceAgentGUARD™
27 named controls in four classes, with the code reference for each.
See the controlsAI Microservices
Ten self-serve tools on Stripe. Pay once, use for twelve months, reports and exports included.
Use a toolThe hard questions
Nine due-diligence questions answered the same way each time: what we claim, what we do not, how you check.
Read the answers
Our history
Arrochar Labs commenced in 2026, built on twenty years of evidence gathered advising CXOs through Arrochar Consulting: organisations in government, aviation, logistics and energy paying six to seven figures a year for platforms that never fitted the work, then paying again to work around them.
In 2026 that evidence became software. An AI-agentic build process produced six products, the MeshGov government platform and a catalogue of twenty-one government services in under a year, at near-zero infrastructure cost. On 1 October 2026 the family was brought together under one name: ASP™, the Arrochar Superintelligence Platform.
Today the company runs as two units. Arrochar Labs builds the software. Arrochar Consulting provides the people, and the company holds AWS AI Partner, Google Cloud AI Partner, Anthropic partner and NVIDIA Inception programme memberships.
The problem we solve
Technology leaders are paying for platforms that do not fit the work, and they cannot yet trust the AI that could replace them.
- Platform spend out of control: six to seven figures a year for monolithic suites, then the same again to configure, integrate and work around them.
- Platforms that do not fit the work: users route around them to spreadsheets and email; every government runs the same twenty-one services and none of the suites were designed for them.
- AI nobody can yet trust: agents can do the platform's job, but a board, an auditor or a regulator cannot yet see what they did or stop them when they go wrong.
- Rogue agents: an agent acting beyond its delegation, on data it was not cleared for, or on an instruction planted in a document.
- Data not ready for AI: no governed glossary, lineage or certified data products underneath the models.
- Decisions without evidence: no tamper-evident record of who decided what, under which delegation, with what reasons.
What the platform lets a technology leader do
- Replace expensive platforms with agentic services built to fit the work.
- Govern every AI they run, on rails whose assurance is reused, not rebuilt.
- Keep every decision with a person who holds the delegation.
- Stop a rogue agent before it acts, and prove it afterwards.
- Run the whole organisation, or a whole government, on one evidence trail.
- Start with one service and add the next without a migration.
Standard rails. Made-to-measure services. Agents prepare. People decide. Every time.
- 35%
- have replaced a ready-made tool with a build
- 78%
- plan to build more this year; 60% outside IT oversight
- 41%
- name security and compliance as the barrier
- US$1b+
- AI governance platform spending by 2030, from US$492m in 2026
- 537
- Australian councils, each under growing pressure to show how it governs its AI
Retool, The Build vs. Buy Shift (2026), 817 respondents; Gartner, 17 February 2026; Australian Local Government Association.
What changed, and when
The buyers now have a deadline, the assessors now have a standard, and the spend is moving. None of that was true two years ago.
- 30 June 2026
Every NSW council was expected to have fit-for-purpose AI governance arrangements in place, after the NSW Audit Office's Local Government 2025 report; every other state and agency reads the same report.
- 2023 to 2026
ISO/IEC 42001:2023 gave assessors the first certifiable AI management system standard, and the EU AI Act's high-risk obligations are phasing in, so a board can now be asked to show its AI governance, not describe it.
ISO/IEC 42001:2023; Regulation (EU) 2024/1689
- 17 February 2026
Gartner put AI governance platform spending at US$492m in 2026, more than doubling to over US$1b by 2030.
- 2026
78% of organisations plan to build more AI tooling this year, 60% of it outside IT oversight, and 41% name security and compliance as the barrier: the demand for governed rails is already inside the organisation.
How it fits together
Not just AI agents. An architecture for accountable ones.
Each part of ASP™ does one job, in order, and hands the next part something it can check. That is the difference between building AI agents and running them so an auditor, a regulator or a minister can see what happened.
Before the first build, Orbit Roadmaps finds where AI creates value and profiles the risk.
After it, Execdive turns what the chain recorded into plain-language insight for executives.
Policy, enforcement, templates, data, controls, evidence. Then services.
- AIG Sentinel defines policy
- Meshbone enforces policy at run time
- Nanolitte carries the controls into reusable service templates
- Databallast controls what data the agent may access
- AgentGUARD™ is the cross-platform control setAcross every step
- Evidence chain records what happened
- MeshGov uses the whole thing for government services
- MeshFin uses the whole thing for banking and financial services
Market potential
Global spend on software as a service each year: the suites and platforms agentic services replace (Gartner forecast for 2025).
AI governance platform spending more than doubles, from US$492m in 2026 to over US$1b by 2030 (Gartner, 17 February 2026).
AI governance and control platforms by 2030: the category AIG Sentinel, Meshbone and AgentGUARD™ sell into, before MeshGov's government operations.
About 600 councils in Australia and New Zealand at the platform list price, before any MeshGov service, build or run fee. Plan: A$1m annual recurring revenue by 2028, A$3m by 2030 (Arrochar Labs estimate).
From MVP to the platform of record for accountable AI
Delivered milestones are live today. The round funds the next four. Nothing after October 2026 is a new product; it is hardening, certification and distribution of what already runs.
- MVPJune 2026
- Six products liveJuly 2026
- Gold Standard AI SafetySeptember 2026
- MeshGov, 21 servicesSeptember 2026
- AgentGUARD™29 September 2026
- Meshbone gateway1 October 2026
- ASP™ and the Orbit app1 October 2026
- MeshFin catalogue5 October 2026
- Industry Agents, 7 packages6 October 2026
- First paying councilQ4 2026 · funded by this round
- ISO/IEC 42001 and SOC 22027 · funded by this round
- AWS production estateH1 2027 · funded by this round
- UK and EU2029 · funded by this round
The round
- Ask
- US$750k on a SAFE
- Valuation cap
- US$3.0m
- Post-money
- US$3.75m, about 20%
- Minimum ticket
- US$25k
What we will take: a US$500k minimum close at a US$2.5m cap, or a US$250k first close from a strategic angel who opens doors. Minimum ticket US$25,000. All amounts in US dollars. Indicative terms; the SAFE instrument governs.
Runway for eighteen months, so the company raises once and executes through revenue.
Built to be bought, run to be independent.
What a SAFE holder receives
- Conversion into the share class the next priced round issues, at the SAFE price set by the US$3.0m post-money cap, with no discount to apply.
- On a sale or listing before conversion, your election: the purchase amount back in cash, or ordinary shares at the SAFE price.
- On an insolvency event, the purchase amount repaid in priority to shareholders, pro rata with other SAFE holders.
- Your entitlement adjusted for any share split or reconstruction, and pre-emptive rights waived by later shareholders in your favour.
- Quarterly accounts from the company, our commitment alongside the instrument.
Clauses 3, 4, 5, 7 and 8 of the SAFE. The signed instrument prevails.
The questions we get asked first
- Who can invest?
- Investors to whom the company can lawfully offer a SAFE without a disclosure document under the Corporations Act 2001 (Cth): sophisticated investors (with an accountant's certificate) and professional investors under section 708. The investor confirms this in the SAFE itself. The offer is made through the documents issued for it, not through this page.
- Is there a minimum ticket?
- Yes: US$25,000. The round targets US$750k, with a US$500k minimum close, and we will take a US$250k first close from a strategic angel who opens doors. All amounts are in US dollars. Larger cheques are welcome; tell us what you have in mind.
- How does the SAFE convert?
- At the next qualifying priced round, into the share class that round issues, at the SAFE price: the US$3.0m post-money valuation cap divided by the fully diluted share count. There is no discount. On a sale or listing before that, you choose cash or ordinary shares. The full mechanics are in clauses 3 and 4 of the SAFE.
- What rights do I have before conversion?
- None as a shareholder: no voting, dividends or pre-emptive rights until shares are issued. You do have priority repayment on insolvency, adjustment for any share split, and our commitment to send quarterly accounts.
- When does the round close?
- It is open for expressions of interest now. We are targeting a close in the first half of 2027, or earlier if the target is reached.
- Is there a data room?
- Yes, on request after the walk-through: the constitution and shareholders agreement, the cap table, IP assignments, the security architecture, the SAFE and the full investor pack. Some of it under a confidentiality undertaking.
- Why a SAFE rather than a priced round?
- Speed and cost at pre-seed. A SAFE lets the round close in weeks on a standard Australian post-money form, and leaves the price to be set properly at the institutional seed we plan for 2027 to 2028.
- What will the money be used for?
- Hardening, certifying and distributing what already runs: the production estate, ISO/IEC 42001 and SOC 2, the first sales and customer-success hire, and the first council deployments, with eighteen months of runway so the company raises once and executes through revenue. The budget is in the investor pack.
Ready to look closer?
Book a walk-through of the full investor pack, or write to investor@arrocharlabs.com, and read the SAFE the round is issued on.
Or write to investor@arrocharlabs.com.
Important information
This page is information about Arrochar Labs. It is not an offer of securities, an invitation to apply for securities, or financial product advice, and it does not take your objectives, financial situation or needs into account.
Any offer of a SAFE will be made by Arrochar Consulting Pty Ltd (ACN 624 705 531), trading as Arrochar Labs, only to investors to whom an offer can lawfully be made without a disclosure document under the Corporations Act 2001 (Cth), such as sophisticated and professional investors under section 708, and only through the documents issued for that offer. The terms on this page are indicative and may change before any instrument is signed.
Plans, targets, milestones and market estimates on this page are forward-looking. They rest on assumptions that may prove wrong, and they may not be achieved. Investing in an early-stage company is high risk: you may lose all of the money you invest, your investment will not be liquid, and a SAFE carries no interest, no maturity date and no right to repayment. Obtain your own legal, tax and financial advice before you invest.
ASP™ and AgentGUARD™ are unregistered marks of Arrochar Labs. Partner programme names and panel names belong to their owners; membership of a programme or listing on a panel is not an endorsement of this round.