Consulting

What Kroy actually does in a consultancy.

  1. It starts with an engagement.

    Scope, stakeholders and deliverables — the engagement lives in Kroy.

  2. Research is gathered.

    Interview notes and market research are attached to the engagement.

  3. Kroy keeps clients apart.

    The AI working on this engagement cannot see any other client’s material.

  4. AI synthesises.

    The AI turns research into a draft recommendation, with sources recorded.

  5. The deliverable is an Artifact.

    Versioned, attributed and approved by the engagement lead.

  6. The client sees the approved version.

    Only the approved deliverable crosses to the client’s projection — not the working notes.

Client engagement
Research
Kroy
AI
Artifact
Client projection
Client engagement
Research
Kroy
AI
Artifact
Client projection
Consulting workflow through Kroy. It starts with an engagement.. Research is gathered.. Kroy keeps clients apart.. AI synthesises.. The deliverable is an Artifact.. The client sees the approved version..

Why consulting

Consultants move between clients, often in the same week. The firm’s value is its accumulated knowledge; its obligation is that one client’s information never informs another’s work without permission. AI makes both halves harder: it is very good at drawing on everything it has seen.

Kroy lets a consultancy give AI the context of one engagement at a time — the client’s documents, the project plan, the firm’s reusable methods — and nothing from the engagement next door.

Engagements as boundaries

Kroy scopes AI access to engagements. An AI acting for a consultant sees the engagements that consultant is staffed on. Firm knowledge — methods, templates, anonymised case material — can be shared across engagements; client-identifiable material stays inside its engagement.

One state of work

Engagements involve many people and, increasingly, several AI tools. Kroy State holds the working record: what the client asked, what was decided, what is outstanding. A senior consultant using Claude and an analyst using ChatGPT pick up from the same place, and a new team member can be briefed from it on day one.

Where AI helps, and where Kroy draws the line

AreaAI canKroy boundary
ProposalsDraft a proposal from the brief and past workOnly reusable firm material, not other clients’ data
DiscoverySummarise interviews and documentsOnly for the user’s engagements
Project deliveryUpdate Jira and prepare status reportsWorkflow transitions need approval
AnalysisWork with the client’s data setsData stays in the engagement
DeliverablesPrepare drafts and slidesClient issue is approved by the engagement lead
Client communicationsDraft updates and repliesDrafts, not sends
CRM and pipelineLog meetings and propose updatesPipeline changes approved by the account owner
CommercialsRates and margins stay with partners

Sharing with clients

Give a client a projection of the engagement — agreed deliverables, open actions, key decisions — and let them use it directly or through their own AI. Internal commentary, pricing and other clients’ material never leave the firm.

AI policy first

Many consultancies advise clients on AI while still settling their own rules. PolicyHQ helps a firm write an AI policy and an access matrix quickly.

Define the rules with PolicyHQ. Put them into practice with Kroy.

Frequently asked questions

Can the AI use what it learned on one client for another? Not through Kroy. Each request is limited to the engagement it is made for, plus the firm knowledge you mark as shareable.

We use Claude and ChatGPT. Do we have to choose? No. Both connect to Kroy, see the same permitted information and work from the same project state.

Can clients see what the AI did on their engagement? You can choose to share a projection of the work with the client. Internally, every AI request is recorded.

Do our system credentials go to the AI provider? No. Kroy holds the connections and returns only the results of authorised requests.

Case Studies for consulting