Skip to content

Platform setup

Platform setup is done once per deployment, not per engagement, by a Platform Administrator. It’s the Admin area (/admin). If your firm’s deployment is already running, most of this is done — skim it so you know where the controls are.

Admin → AI Quality. Each AI function (extraction, scoring, and so on) has a precision and a recall threshold. These govern how confident the AI has to be before its output is treated as reliable, and they feed the risk-tiering of extracted candidates. The defaults are sensible; change them only with a reason.

Knowledge (/knowledge). Your firm’s curated body of standards, patterns, and playbooks — see The standards library for the full picture. A fresh deployment starts empty; you build it up by uploading standards and by promoting what works at the end of each engagement.

Admin → User Management. Create accounts, deactivate them, reset access. Engagement-level roles are assigned per engagement on its team page — this screen is just the account list.

Which domain packs are active is a deployment setting (an environment variable), not a screen. See Domain packs. Talk to whoever runs the deployment if you need a different pack active.

Admin → Metamodel. The deployment-wide rulebook of entity types, relationships, scoring dimensions, and viewpoints — disable or rename what your firm doesn’t use, or define your own. See The metamodel. An Engagement Lead can overlay further for a single engagement without touching this firm-wide rulebook.

  • Client Management — the register of end clients (each is a separate tenant).
  • Platform Overview / LLM Usage — dashboards: engagements in flight, AI spend by provider.
  • Closure Requests / Gap Unlock / Score Unlock Requests — approval queues for things a consultant can request but only an admin can grant.
  • SKB Nominations — standards a consultant proposed for the library, awaiting your approval.
  • Data Maintenance / Reconciliation — housekeeping for orphaned data after a deletion.