Getting started
What it takes to begin, honestly.
The engineering is about a day. Everything that takes longer than that is yours, and no vendor can shorten it. So we start those parts first and say so, rather than quoting you the day and letting you discover the rest.
The fast part
About a day of our time.
An address
Your own hostname for the connector, with a certificate. Nothing is installed on anybody's machine.
A credential in a vault
One per environment, held so that no person and no tool ever sees it.
The people and their grants
Who may reach which environment, with which tools. Usually three or four people to start, not everybody.
Your written rules, drafted
We read what you have connected and propose the vocabulary and conventions. A named person at your company approves it before anything uses it.
The slow part
Yours, and genuinely not shortenable.
Both of these have run to weeks at organisations of the size this is built for. Neither is a reason not to start; both are a reason to start them now rather than after a pilot.
| What | Who owns it | Realistic |
|---|---|---|
| Your security review | Your security team | Weeks. Everything they will ask for is published rather than gathered on request, which is the only part we can make faster. |
| Directory consent | Your identity administrators | Depends entirely on your organisation. At larger companies a third-party application signing in against your directory needs an admin consent process. Assume it needs one. |
Directory consent is the long pole, not the engineering. If you take one thing from this page: test a single work-account sign-in against your directory early, before anybody agrees a date for anything else.
The shape of a start
One system, a few people, a fortnight of rows.
- 01
Pick the system where a wrong answer would hurt
Not the easiest one to connect. The value shows up where the questions are real.
- 02
Three or four named people
Enough to produce real usage, few enough that the grants are obvious and reviewable.
- 03
Read the audit trail after two weeks
This is the actual evaluation. What people asked, what was refused, and whether the refusals were right.
- 04
Decide with evidence
Widen, narrow, or stop. All three are reasonable outcomes and the rows will tell you which.
Start with the review, not the demo.
Send your reviewer the pack first. If it does not survive them, a demo was never going to help.