KyperX AI / Tools
Free tools
Free AI planning, governance and infrastructure tools.
Seven tools. No email address, no account, nothing uploaded.
Each one covers a single decision and produces something you can take into a meeting: a business case, a register, a pilot charter, a procurement pack, a decision record, a cost comparison or a sizing worksheet. They start empty, run in your browser, and you can load an example to see the shape before you enter anything.
Two tools
Plan an AI opportunity.
Quantify a candidate use case in the organisation's own figures, then turn the pilot into something that can be tested rather than asserted.
About 10 minutes
Estimate annual benefit and payback
Work out what one AI use case is worth, keeping hours released and money saved as separate claims.
- For
- Whoever has to put a number in front of a finance reviewer
- You need
- Process volume, handling time, a loaded hourly rate, and what it would cost to build and run
- You get
- A one-page business case: hours released, benefit, first-year net, payback, and every assumption beside it
About 12 minutes
Write a pilot you can actually judge
Fix the baseline, the target and the stop condition before the pilot starts, not after the argument.
- For
- Whoever owns a pilot that has to reach a decision
- You need
- The baseline you measure today, the target you are willing to state, and three review dates
- You get
- A printable pilot charter: baseline, acceptance criteria, measurement schedule, rollback triggers and the decision questions
Three tools
Govern and procure AI.
Find out what is already in use, decide what a supplier must satisfy, and establish which deployment patterns your own constraints leave open.
About 5 minutes per system
Build an AI register and see what needs a decision
Turn what people tell you into a register, with the gaps named rather than guessed.
- For
- Whoever has been asked what AI the organisation is using
- You need
- What each tool is, who owns it, what data reaches it, and who holds the account
- You get
- An AI systems register, a summary, flags explained against published conditions, and a list of what nobody could answer
About 8 minutes
Generate the requirements and questions for an AI purchase
Separate what a supplier must satisfy from what you would prefer, before anyone writes a response.
- For
- Whoever is about to buy an AI product
- You need
- The outcome you want, the data involved, and how much control you need over deployment and exit
- You get
- A printable pack: gates, requirements, evidence requests, a grouped questionnaire, acceptance criteria and a matrix you weight
About 6 minutes
Find out where this is allowed to run
Six deployment patterns, screened against your own constraints by rules printed on the page.
- For
- Whoever has to choose between hosted, private and owned
- You need
- Your sensitivity, residency, retention, network, latency and control position
- You get
- A decision record: which of six patterns remain, why each survived or did not, and the evidence to request
Two tools
Size private AI infrastructure.
Two worksheets for an owned or dedicated deployment: what it costs against hosted API pricing, and what it takes to serve the load.
About 10 minutes
Compare running your own model against paying per token
Owned against hosted on your own figures, with the cash view kept separate from the annualised one.
- For
- Whoever has to justify owning infrastructure
- You need
- Token volumes, the API prices quoted to you, hardware cost, power and who would operate it
- You get
- Annual owned cost against annual API cost, a first-year cash figure, a break-even volume and three scenarios
About 8 minutes
Size the GPUs, power and cooling
A production sizing worksheet that asks for measured performance instead of inferring it.
- For
- Whoever has to specify hardware for a deployment
- You need
- Peak demand, and throughput and concurrency figures measured on the stack you intend to run
- You get
- GPU and server counts, the binding constraint, IT load, facility input, heat and monthly energy
Which one
Pick by the question you arrived with.
| If you are asking | Use | Time | You need | You leave with |
|---|---|---|---|---|
| Is this use case worth doing, and what would I put in front of a finance reviewer? | Estimate annual benefit and payback | About 10 minutes | Process volume, handling time, a loaded hourly rate, and what it would cost to build and run | A one-page business case: hours released, benefit, first-year net, payback, and every assumption beside it |
| We are about to run a pilot. How will we know whether it worked? | Write a pilot you can actually judge | About 12 minutes | The baseline you measure today, the target you are willing to state, and three review dates | A printable pilot charter: baseline, acceptance criteria, measurement schedule, rollback triggers and the decision questions |
| What AI is already being used here, by whom, on what data? | Build an AI register and see what needs a decision | About 5 minutes per system | What each tool is, who owns it, what data reaches it, and who holds the account | An AI systems register, a summary, flags explained against published conditions, and a list of what nobody could answer |
| We are going to buy an AI product. What must the supplier satisfy, and what do we ask? | Generate the requirements and questions for an AI purchase | About 8 minutes | The outcome you want, the data involved, and how much control you need over deployment and exit | A printable pack: gates, requirements, evidence requests, a grouped questionnaire, acceptance criteria and a matrix you weight |
| Given our constraints, where is this allowed to run at all? | Find out where this is allowed to run | About 6 minutes | Your sensitivity, residency, retention, network, latency and control position | A decision record: which of six patterns remain, why each survived or did not, and the evidence to request |
| Is running our own model cheaper than paying per token? | Compare running your own model against paying per token | About 10 minutes | Token volumes, the API prices quoted to you, hardware cost, power and who would operate it | Annual owned cost against annual API cost, a first-year cash figure, a break-even volume and three scenarios |
| How many GPUs, how much power, and how much heat? | Size the GPUs, power and cooling | About 8 minutes | Peak demand, and throughput and concurrency figures measured on the stack you intend to run | GPU and server counts, the binding constraint, IT load, facility input, heat and monthly energy |
How they work
The same shape on every page.
What you do
- Fill in the fields marked required. Optional ones are labelled as such.
- Press Generate result. Nothing is calculated until you do, and an out-of-range value is refused rather than computed.
- Copy, download a CSV, or print the artefact on its own. All three are built in your browser.
- Open Try an example first if you would rather see a filled-in version. The figures in it are invented and labelled.
What they will not do
- Score you. No readiness percentage, no maturity level, no percentile — a score needs a benchmark population, and inventing one would be worse than having none.
- Tell you anything is compliant, secure, safe or approved.
- Carry a vendor price list. Where a commercial figure is needed you enter it, in a currency you name.
- Attribute a result to KyperX. Every figure is arithmetic on numbers you supplied.
Your inputs stay in your browser
Each tool is one static page plus one same-origin script. There is no form submission, no API and no third-party request, and nothing you type reaches an analytics call. Copy, CSV and print are assembled from what is already rendered.
Nothing is written to browser storage either, which has one consequence worth knowing: reloading clears what you entered. Export before you leave. The shadow AI register can read its own CSV back in, and that file is read in your browser rather than uploaded.
The site records anonymous, cookieless page analytics. Each tool additionally records that it was opened, that it produced a result, that an export was taken and that a link was followed — two fields, which tool and which of those four things, and never a value or a count. Our privacy policy sets out what the site records in full.
Common questions
Are these free, and what is the catch?
Free, with no gate. If a tool is useful, some people who use it will later want the work it points at. That is the whole mechanism, and it is why none of them asks for an email address.
Is anything I type sent to KyperX?
No. Every calculation happens in your browser, there is no server to receive it, and the copy, CSV and print functions build their output from what is already on the page.
Do they work without JavaScript?
The input controls do not, and they are hidden rather than shown as dead boxes. Everything else does: each page carries a worked example computed when the site is built, the formulas and decision rules, a reference table, the limits and the definitions.
Can I use the output in a board paper or a tender?
Yes. Print emits the generated artefact on its own, and the CSV states what it assumed. Read the limits for the tool you used, and treat the governance and procurement outputs as a starting point for your own drafting.
Why is there no AI readiness score?
A score is only meaningful against a population, and we have no defensible population to compare you with. Publishing one would mean inventing the benchmark. These tools make your own position explicit instead.
Something is wrong. How do I tell you?
Email ai@kyperx.com with the page and the figure or rule you think is wrong. Please do not send the data you were working from; we do not want it and cannot receive it through the tool.
Primary sources used across the governance tools
Cited on the individual pages next to the point they support, with review dates, because they change.
- National AI Centre, AI systems register. Register template published 22 April 2026.
- National AI Centre, Guidance for AI adoption: foundations. Part of the Essential AI practices.
- National AI Centre, Guidance for AI adoption: implementation guidance. Part of the Essential AI practices.
- Office of the Australian Information Commissioner, APP Guidelines chapter 8: APP 8 cross-border disclosure of personal information. Updated 3 October 2025.
- Office of the Australian Information Commissioner, Guidance on privacy and the use of commercially available AI products. Published 21 October 2024, updated 17 January 2025.
- Digital Transformation Agency, Guidance on AI procurement in government (announcement and links). Released 2 December 2025; the guidance itself is hosted on BuyICT.
- Digital Transformation Agency, Technical standard for government's use of artificial intelligence. Applies to Australian Government agencies.
Primary sources last reviewed 27 September 2026. These are the publishers’ own pages; they are quoted nowhere on this site and nothing here restates them as a requirement that applies to you.
If a tool has told you something you want to act on.
The engagements these point at are published with starting prices. An enquiry is a conversation about scope, not a commitment.
Australian organisations: engagements and pricing in AUD.