01
Half price for the first three months
—
Conditional on the application opening at all. No card is collected now.
Record AR-REG-001 · Founding Researcher Register
Register as a founding researcher.
Then read the small print.
The register records who is interested in the planned AlphaRail Foundry application, what they trade, and — the part that actually matters — which single module they would need first. It buys you no place in a queue and no discount on a price that does not exist. What it does buy is influence over the order in which things get built, and that is worth more than it sounds.
Read this before filling anything in
The form on this page is a demonstration. It validates your input in your own browser and shows a confirmation, and that is the whole of what happens. Nothing is transmitted, nothing is stored, no email is sent, no address is added to a list, and no record survives you closing the tab. If you want to be contacted, use the contact page instead.
Founding Researcher · AR-REG-001·OFFER
Half price for your first three months,
on whichever plan you choose.
—
01What a Founding Researcher GetsAR-REG-001·01
Six specific things,
and deliberately no more.
This list is short and modest on purpose. A long list of benefits attached to a product that does not exist is a list of things nobody can deliver. Each item below is something that could be provided without the application being finished, which is the test each one had to pass — including the discount, which costs nothing to honour if the application never opens, because there would be nothing to charge for either.
02
Early access to modules as they become usable
Not the whole application — individual modules, at the point where one becomes operable rather than at the point where everything is finished. That means seeing a rough version of the data intake or the comparison arena before it is good, with the parts that do not work still visibly not working.
Conditional on modules reaching that state at all.
03
Influence over the build order
The dependency chain is fixed — data before simulation, simulation before diagnosis, diagnosis before governance — but within each stage there is real latitude. Which of the five Diagnose modules is built first is a genuinely open question, and the answer should come from people who would use it.
The most substantial of the four.
04
Research notes as they are written
The working notes behind the archive records: what was tested, what failed, which assumptions in the preview engine turned out to be indefensible and were changed. Unpolished, occasionally wrong, and considerably more useful than a finished summary.
Written anyway; sharing them costs nothing.
05
A weekly letter while it is being built
One letter a week: what was built that week, what broke, what was decided and why. It carries the design architecture as it settles — the scoring hierarchy, the parameter catalog, the zone inventory — and the functionality and feature notes ahead of release, including the roadmap for the granular Monte Carlo controls. It stops being weekly when there is nothing worth sending, rather than being padded to hit a schedule.
Unsubscribing is one click and does not forfeit the founding discount.
06
A demonstration built from your answers
Once the application is fully deployed, anyone who completed the questionnaire below can request a demonstration built around what they answered — their market, their strategy style, their trading experience, their preferred interface depth and the module they said they would need first. A macro trader asking about regime dependency and a scalper asking about friction should not be shown the same forty minutes.
Conditional on deployment. It is not a sales call disguised as a demo.
What is deliberately absent from that list
No lifetime licence, no rate locked in beyond the three months, no priority position, no exclusive strategies, no private community, no course, no calls. Every one of those is a common early-access inducement, and every one of them would be a promise about a future that has not been secured. The half-price period is deliberately the narrowest version of that kind of term: a discount on a stated price for a stated number of months, ending on a date you can calculate yourself.
02What Is Expected In ReturnAR-REG-001·02
Feedback, and
a specific kind of honesty.
The word researcher in “founding researcher” is meant literally. This is not an audience being courted ahead of a launch; it is a small group of people being asked to do the thing this site argues for — examine something carefully and say plainly what does not survive the examination.
Tell us what does not workSpecifically. “The regime page is confusing” is a starting point; “the transition matrix implies a persistence that the timeline contradicts” is something that can be fixed this week.
Argue with the modelThe preview engine encodes claims — that friction scales against the stop, that partials trade tail for capture, that correlation converts concurrency into concentration. Those claims are arguable and should be argued with.
Say when a feature is not worth buildingMore useful than requesting one. A module cut before it is built is the cheapest possible improvement, and the Studio tier in particular has never been tested against anybody's real requirements.
Report anything that reads as a promiseIf a page implies a guaranteed result, an assured launch, or that a Level 2 preview is doing real work, that is a defect. It should be reported like any other.
Accept unfinished softwareEarly modules will be rough. The exchange is honesty in both directions: nothing is presented as more finished than it is, and nothing is expected to be judged as though it were.
What is not expected
No testimonial, no review, no referral, no promotion, no case study, and no trading history. You are never asked to share account statements, broker credentials or performance figures — not now, and not in any planned version of the application. Nothing on this site is designed around collecting them.
You are also not expected to remain interested. A registration that goes quiet is a perfectly reasonable outcome and carries no obligation whatsoever — particularly since, in this phase, the registration is not recorded anywhere in the first place.
No testimonial
No trading history
No obligation
03Explicit Non-PromisesAR-REG-001·03
Five things this registration
explicitly does not do.
Stated before the form rather than after it, because a disclosure that appears only once somebody has already submitted their details is not a disclosure.
NOT 01
It does not reserve a place
There is no queue, no position, no waiting number and no ordering by date of registration. If a limited group is ever invited to an early module, it will be chosen for the mix of markets, experience levels and use cases it represents.
Most registrations will not be invited
NOT 02
It does not fix a price
No founder rate, no early-bird discount, no lifetime deal, no grandfathering. The tier proposal has no prices in it at all, and offering a discount on an undecided figure would be meaningless.
See the tier proposal
NOT 03
It does not commit to a date
The largest remaining piece of work — accounts, storage and versioning — has not started. Every module downstream inherits that uncertainty, and the roadmap deliberately orders the work without attaching dates to any of it.
No quarter, no year, no “soon”
NOT 04
It does not promise the application at all
Nineteen of twenty-five modules do not exist. The project may stop at the preview stage, and in that case there is no partial product, no reduced offer and no fallback tier.
A real possibility, not a formality
NOT 05
It does not improve your trading
No tool described here identifies an edge, guarantees that one persists, or protects an account from a market that changes behaviour. A better instrument produces better questions; it does not produce a better result on its own.
No performance claim, at any tier
And one about this phase specifically
Because the form transmits nothing, there is currently no register at all — not an empty one, not a local one, not a file. Submitting the form produces a confirmation message and a reference string generated in your browser for demonstration purposes. If a real register is ever created, this page will say so plainly and the disclosure above will change.
If you want an actual reply
Use the contact page. It explains what correspondence is possible in this phase and what is not, and it does not pretend that a form submission reaches anybody.
Contact →04Methodology · The Form ItselfAR-REG-001·04
What happens
when you press submit.
The site's honesty rules apply to its own registration form as strictly as they apply to a chart. Here is the exact sequence, with nothing omitted.
Step 01 · Validation
Each required field is checked in your browser. Empty required fields, an email address that does not have the shape of one, and an unticked consent box each produce a message directly beneath the field, and the first failing field receives keyboard focus. No network request is involved.
Step 02 · Confirmation
If every check passes, a confirmation appears with a reference string of the form AR-REG-nnnn. That string is generated locally for demonstration purposes and corresponds to no record anywhere. The form then clears itself.
Step 03 · Nothing else
No request is sent, no cookie is written, no analytics event is fired, no third-party script runs, and nothing is placed in local storage. The site as a whole makes no external requests at all — the fonts, the styles and every script are served from the same origin.
Demonstration form — nothing is transmitted or stored
05RegistrationAR-REG-001·05
The register.
Nine questions and one confirmation. The two that carry the most weight are the module you would need first and the free-text box at the end, because those are the only ones capable of changing a decision. Everything else is context for reading them.
50%
—
—
Demonstration form — nothing is transmitted or stored
What exists today
Every one of those working modules can be operated right now, without registering, without an account, and without anything being stored. Registration is about what comes after them: being told when a module becomes usable, and half price for the first three months on whichever plan you choose when the application opens.
The module register →Data collected: none
No analytics, no cookies, no tracking pixels, no third-party requests, no local storage. This is not a policy statement about intentions — it is a description of what the code on this page does.
06Current Build OrderAR-REG-001·06
What is being worked on,
in the order it is being worked on.
This is the list your registration is capable of changing. The dependency constraints are fixed, but the sequence within them is not, and a well-argued case for moving something up is exactly the kind of feedback the register exists to collect.
| # | Item | State | Why it sits here | Movable? |
|---|
What would actually move something
A concrete account of the work you are trying to do and the point at which the current tools stop helping. “I keep four branch populations in one spreadsheet and cannot see which one is deteriorating” is an argument. “Please add optimisation” is not, and optimisation is the item most likely to be requested and least likely to help.
What cannot move
The persistence layer stays first regardless of how many people ask for something else, because thirteen of the twenty-five modules are blocked behind it. Reordering around it would produce features that cannot keep their own results.
07Questions About RegisteringAR-REG-001·07
Six questions
about the register.
The subset relevant to this page. Everything else — the preview engine, the archive, accessibility, the MARS boundary — is answered in full on the questions page.
In this phase, the point is to see the form work and to read the questions, which are themselves a
statement about what the project thinks matters. The questions are the deliverable as much as the
answers are. If a real register is ever built, the disclosure on this page changes first and the
form changes second — not the other way round.
No. There is no mailing list in operation, no email service connected to this site, and no address
is captured — the field is validated for shape and then discarded when the form clears. Nothing
you enter here will produce a message, now or later.
No. The two products are separate, and the register is not a sales channel for either. The MARS
question exists because a current workbook user has a different perspective on the governance
modules than somebody meeting the ideas for the first time, and both perspectives are worth having.
Answering “not familiar” costs you nothing.
The list is the complete application register — all twenty-five modules across the five stages — so
if what you need is not there, it is genuinely not planned. That is worth knowing and worth saying:
use the comments box. A missing module is more informative than a popular one.
No. Nothing on this site can be purchased, there is no checkout, and no tier has a price attached
to it. Taking money for access to something that does not exist would be the single clearest
contradiction of everything the rest of the site argues for.
Nothing at all. The Foundry does not give advice, does not evaluate your account, does not know
what you trade beyond the two context questions above, and makes no claim that using it changes an
outcome. The risk disclosure states the position in full and applies to every page on this site.
08LimitationsAR-REG-001·08
The limits of
an early register.
Registers of this kind have well-known failure modes. Naming them is cheaper than pretending they do not apply here.
Interest is not demand
People who register for something free and unbuilt are a poor predictor of people who would use something priced and finished. Whatever this register eventually contains should be read with that discount applied.
The sample selects itself
Anybody who reaches this page has already read several thousand words of technical argument. That is a specific and unrepresentative population, and building only for it is a way of narrowing a product without noticing.
Stated preference is weak evidence
The module somebody names in a dropdown is frequently not the one they open first. It is still the best signal available before anything exists to observe, which is why the question is asked and why the answer is weighted modestly.
The form itself proves nothing
A working validated form is a Level 1 feature of a static site. It is not evidence of a backend, an account system, or any of the infrastructure the application would need, and it should not be read as a sign of progress toward those.
Nothing here is contractual
The register creates no agreement, no entitlement and no relationship. The terms page governs use of the site; this page adds nothing to it and takes nothing away.
The project may simply stop
The most likely single outcome for any unbuilt software project is that it is not built. That possibility is not hedged, softened or hidden behind an optimistic roadmap anywhere on this site.
If you would rather just read.
Registering is optional and always will be. The most valuable material on this site is the part that has nothing to do with the application: fourteen learning modules on the measurements traders routinely misread, and an archive of research records that keeps its failures alongside its successes. Neither requires an account, an email address, or any interest in a future product.
Start here if you want to learn
Expectancy as a rate rather than a promise, drawdown as a path rather than a number, and why the same trades in a worse order produce a different account.
Learning centreStart here if you want the evidence
Every record with its method, abstract, status and verdict — including the regime study that failed precisely where it would have been useful.
Research archiveNo account required · nothing is uploaded · every output is a labelled model result