New hires ramp slowly when the knowledge they need lives in people's heads. We map what the role must know and turn it into an onboarding path. AI surfaces gaps and answers ramp questions from your knowledge base; a person owns and verifies the path.
Common pain points
What the role actually needs to know was never written down, so a new person learns by asking the same questions for weeks.
Knowledge is tribal — one or two people hold the working context — so when they are away or busy, ramp stops.
When someone leaves the role, their knowledge leaves with them; the next person rebuilds it from scratch.
Onboarding is a list of meetings and tools, not a path that ends at a clearly capable owner of the role.
The knowledge base has answers somewhere, but a new person does not know what to read first or what is current.
Nobody owns the onboarding path itself, so the gaps a previous hire hit are still there for the next one.
The system has three parts. A role knowledge map — what the role must actually know. An onboarding path — a sequence with a named owner per step. And a gap list — the tribal knowledge to capture before it walks out. AI assembles the path from existing docs, drafts the knowledge map, flags gaps, and answers ramp questions from your knowledge base. A person owns the path and gives the role-ready sign-off; AI never certifies competence.
How we work
We define what the role must know, with the people who already do it, and write the role knowledge map and the onboarding path with a named owner per step.
We connect AI to assemble the path from existing docs, draft the knowledge map, flag the gaps a new person will hit, and answer ramp questions from the knowledge base.
We name who verifies the path is complete and the role-owner sign-off, and hand over a runbook the team operates and edits.
What we build
Area 1
We write what a person in the role must actually know to operate — the systems, the decisions, the people, the recurring exceptions — agreed with the people who already do the role. It is the answer to "what does this role need in their head?" written down.
Deliverable
A role knowledge map: the topics, systems, decisions, and people a person in the role must know to operate.
Area 2
We design the sequence a new person follows — what to read, who to meet, what to do hands-on, in what order — with a named owner per step who is responsible for that step landing. It ends at a clearly capable owner of the role, not a checklist of meetings.
Deliverable
An onboarding path: a sequence of steps with a named owner per step, ending at role-ready sign-off.
Area 3
We surface what only one or two people know about the role and flag it as a gap to capture — before the next departure makes it lost knowledge. The gap list is what turns tribal knowledge into a queue the team can actually work through.
Deliverable
A gap list: the tribal knowledge items to capture, who holds them today, and the order in which they get written down.
Area 4
We connect AI to assemble the onboarding path from existing docs, draft the role knowledge map, flag gaps a new person will hit, and answer ramp questions from the knowledge base. What AI may never do — certify the new person is ready or replace the role-owner sign-off — is written down. AI prepares; a person verifies.
Deliverable
An AI capability map with the limits written: what AI assembles/drafts/flags/answers, and what only a person verifies.
Area 5
We wire AI to answer the common ramp questions from your existing knowledge base — with sources cited — so a new person stops blocking on the one person who knows. Where the knowledge is missing, the answer says so honestly and feeds the gap list.
Deliverable
A ramp Q&A: a path for a new person to ask and get answers from the knowledge base, with sources and gap-flagging where the answer is missing.
Area 6
We hand the system over as a runbook: the role knowledge map, the onboarding path, the gap list, the AI limits, and the verification step. The goal is for the next hire — and the one after that — to use a sharper version, not start from scratch.
Deliverable
A runbook with the knowledge map, onboarding path, gap list, AI limits, and verification step the team owns.
What the role must know is written down, so a new person stops learning by asking the same questions for weeks.
The onboarding path has named owners per step, so each piece actually lands and ramp does not stall when one person is busy.
Tribal knowledge is on a queue to be captured, so a departure stops being a loss event for the team.
A new person can ask the knowledge base and get answers, so they stop blocking on the one person who knows.
AI assembles, drafts, flags, and answers — and a person verifies, with the limit that AI never certifies competence written down.
The path lives in a runbook the team operates and edits, so the next hire arrives to a sharper version of the same system.
Internal knowledge-base design is about the structure of your KB — what areas exist, what lives where, who owns each area. This is about transferring to a new person in a role. We use your KB if you have one; we do not design it. If your real problem is that your KB has no shape, that is the KB-design work, and we will tell you.
Internal knowledge operations is about keeping the KB alive — findable, owned, kept current. This is about transferring to a new person in a role. We feed off a well-run KB; we do not run it. If the KB is decaying, that is the knowledge-ops work and a healthier inbound to this.
AI-assisted internal search routes a person to what the team already knows. This is about turning that knowledge into a path a new person can follow to become capable in a role. Search is one piece of the ramp Q&A; this is the whole onboarding system.
No. We do not sell or build an HRIS, a recruiting tool, or a course/LMS platform. The deliverables are the role knowledge map, the onboarding path, the gap list, the AI ramp Q&A, and the verification — set up in the tools your team already uses. We name current tools as examples; we do not sell or integrate any of them.
No. AI assembles the path, drafts the knowledge map, flags gaps, and answers ramp questions — but a person owns the path and the role-owner sign-off. The rule that AI never certifies competence is written into the system, not left to default.
It does not include designing or operating your knowledge base (those are separate pages), an HR/HRIS platform, a recruiting tool, a course/LMS, or an AI that certifies the new person is ready. If what you need is one of those, we will tell you and point you to the right work.
Work gets lost moving between teams: sales to ops, ops to support. We define what must travel, who confirms receipt, and when the handoff is ready. AI checks…
Read moreKnowledge ops, not a platformWe help your B2B team organise the knowledge you already have, make it findable, assign clear owners, and keep it updated — using the tools your team already…
Read moreStructure and ownership, not a platformWhen your knowledge base has no design, people stop trusting it. We design the structure — what areas exist, who owns each, what gets archived — inside the t…
Read moreBook a 30-minute call to look at how a role ramps today. We will talk through what the role must know, where the knowledge lives, who holds the tribal bits, and where AI can assemble the path and answer ramp questions. If the real issue is the knowledge base itself, we will tell you and point you to the work that fits.
Book a 30-min call