Is it ready for real customers?
A review of one agent or one workflow, before it touches real customers, money or data. We check what your AI can actually do against what it should do, and give you a straight answer: go, go with changes, or not yet.
$1,200set price, incl. GST
10 business daysonce we confirm we can help
One agent or workflowper review
Six multiple-choice questions, about two minutes. You get an answer straight away: a good fit, worth a real conversation, or not a fit right now. We only call if you ask us to.
Founders and small business owners, just before it goes live.
The right time is before your AI starts dealing with customers, their money or their data.
- Founders who built, or are building, with tools like Claude Code, Cursor or Lovable, and are about to launch, take payments or hand the product to real users.
- Small business owners about to switch an AI agent on, or who already have one running and were surprised by something it did.
Built it yourself, or never opened a terminal? Either works. Every review comes with a plain-English summary for whoever makes the decisions, and a technical appendix for whoever builds and maintains the system.
A checked answer, with the paperwork to back it.
Anyone can get a plan written. What you are paying for is someone checking the plan against the system that is actually running.
- A verified boundary check. What your agent can actually do, checked against what it should do: its permissions, keys, data access, what leaves your system, and which of its actions cannot be undone.
- Fixed-rule checks. Automated checks that either match or do not, reported separately from our own judgment so you can tell which is which.
- A risk register ranked by severity. Every finding we make, with what could happen and how to fix it. We don’t leave anything out to make the list look shorter.
- Findings mapped to Australia’s six essential AI practices, from the National AI Centre’s Guidance for AI Adoption, so you have an outside benchmark rather than only ours.
- An owner summary and a technical appendix. The same findings, written for the person who decides and the person who builds.
- A written verdict: Go, Go with changes, or Not yet. With the reasons, and Virasai standing behind it.
Five steps, paid in three parts.
All prices include GST, and you get a tax invoice for every payment.
- Reserve your review ($240 deposit). We confirm your start week by email and send you our intake documents.
- Intake and kickoff ($480). Tell us about the agent or workflow, the systems it touches, and what going wrong would actually cost you. Fill in the intake documents by email, or on a video call if you prefer; calls usually take 30 to 60 minutes. Bring a Role Card if you have one; it makes this part faster.
- Can we help? Within 48 hours of getting access to your project, we tell you in writing whether we can.
- The review (10 business days). We check the real system against what it should be doing, run the fixed-rule checks, and write up every finding.
- Verdict and walkthrough ($480). Your report is released on the final payment. We go through the findings and the verdict with you, by email or video call, whichever you prefer.
If we can’t help, you get back everything you’ve paid. The review also stands on its own: if you don’t need us for anything after it, we will tell you so.
We use the same tools. The difference is what we check them against.
- An independent reviewer. Grading your own work with your own AI shares your blind spots. You only ask about the risks you already know exist.
- We check the system that’s running. Live permissions, keys, data access and what your agent can actually reach. A description of what it’s supposed to do can’t show you those.
- Fixed rules do the checking. A rule either matches or it doesn’t, so there’s no second model’s opinion to talk round. Watch shows the same method running in public on open-source packages.
- Willing to say “not yet”. AI tends to agree with whoever is asking. We will tell you if it is not ready, or not worth doing.
- We use the same tools. Frontier models are part of how we work every day. The difference is what we check them against.
Common questions.
What access do you need?
Enough to see what your agent can actually reach: its configuration, the permissions and keys it runs with, and the systems it connects to. The intake documents include a short access guide, and we use the least access needed, read-only wherever possible. If you would rather not give access at all, you can export settings and logs for us instead. When the review ends, we remove our access, or ask you in writing to revoke it.
Is our data safe?
We keep your information confidential and use it only for your review. Anything raw, like exports or logs that may contain your customers’ information, is deleted when the review ends. We keep an encrypted record of the engagement (your intake documents, our findings and the report) so we can support you if you come back, and we delete that too if you ask.
What if we don’t have an agent yet?
Then a review has nothing to check yet. Start with the free Role Card: it helps you write down the job first, and it becomes the starting point for a review later. Get the Role Card.
What if it turns out bigger than one agent?
The review covers one agent or one workflow, and we agree which one before it starts. If we find something that needs a bigger look, we tell you in the verdict and you decide what, if anything, happens next.
Tell us which agent or workflow.
$1,200 incl. GST, 10 business days, one agent or one workflow. Everything after the review is priced from what it finds, so you never pay for guesswork. See Pricing for the full picture.
Six multiple-choice questions, about two minutes. You get an answer straight away: a good fit, worth a real conversation, or not a fit right now. We only call if you ask us to.