How private beta access opens
Quant Compose is in private beta. Access opens in stages by cohort, invitations arrive by email, and there is no self-serve sign-up during this period.
Quant Compose is currently in private beta, and it is worth being direct about what that means in practice, rather than leaving it as a badge on a page. There is no self-serve sign-up during this period. Registration on the hosted workspace is set to closed by default, which means an account cannot be created by simply arriving at the sign-in page and filling in a form — every account during private beta begins with an invitation.
How access actually opens
Access opens in stages, by cohort, rather than all at once to everyone who has expressed interest. This is a deliberate operational choice, not a technical limitation dressed up as one: bringing in a smaller group of researchers at a time makes it possible to actually read what each cohort reports back, rather than accumulating a backlog of feedback nobody has the attention to act on properly. An applicant who submits an interest form is not given an account on submission — they are added to a queue that is worked through cohort by cohort, and each cohort’s invitations go out by email once a place has opened for them.
There is no public timeline promising a specific date a given applicant will be invited, because the pace of opening cohorts depends on how the current one is going — how much the workspace needs to absorb from what beta researchers report before it is ready to support more people at once. An applicant hears nothing further from the workspace until they receive that invitation email, at which point the email itself carries the steps to create an account and get started.
What an invitation actually grants
An invitation link is a single-use, hashed credential tied to a specific email address — it is not a generic sign-up code that could be shared or reused, and every registration mode in production requires exactly this kind of invitation from an existing account owner. Once used, an invited researcher gets a full research project: an isolated universe, dataset, model and strategy environment with its own audit trail, able to move through the workspace’s full eight-stage research journey from universe selection through to an exported evidence report.
What beta researchers are asked to do
An invitation to the private beta is not simply access granted with no expectation attached. Beta researchers are specifically asked to use the workspace on real research questions of their own — not a scripted demo scenario, but an actual question they would otherwise have tried to answer some other way, so that what comes back reflects genuine use rather than a guided tour.
They are also asked to report refusals and weak results, not only successes. A workspace that refuses to compute something — because a required column is missing, because a stress-test’s historical window is not covered by the loaded data, because a validation split’s evidence pack came back with a section left empty for lack of the underlying result — is doing exactly what it is designed to do, and a beta researcher’s report of that refusal, including whether the stated reason was clear enough to act on, is more useful feedback than a report that everything worked smoothly. The same goes for a result that came back weak or unconvincing: a low deflated Sharpe ratio, a validation grade that failed to pass, a worst fold close to the naive baseline. These are not bugs to be quietly worked around; they are the product doing its job, and knowing whether the explanation that came with them was clear is exactly the kind of structured feedback the beta period exists to gather.
Finally, beta researchers are asked to keep a research log of their own — a record, in whatever form suits them, of what they tried, what the workspace told them, and where they got stuck or surprised. This is a genuinely useful habit in its own right for anyone doing quantitative research, and for the beta period specifically it gives a beta researcher’s later feedback a concrete trail to refer back to, rather than relying on memory of a session from several weeks earlier.
Why this shape, rather than an open launch
An open, self-serve launch would bring in far more people far faster, and it would also make it much harder to give any individual piece of feedback the attention it deserves before deciding what to change. The private beta’s staged, invitation-based structure trades speed of growth for depth of feedback during the period where the workspace most needs that depth — before its research pipeline, its validation methodology and its evidence reporting are locked into patterns that would be expensive to revisit later. Applicants who have not yet received an invitation are not being deprioritised arbitrarily; they are waiting for the next cohort a smaller, more attentively supported one has made room for.