Fabric Readiness
Sprint
A fixed-scope engagement to find the model, logic, and governance gaps stopping Power BI, Fabric, and AI workflows from producing trustworthy answers.
If Copilot sounds great but
your data isn't ready for it.
You've invested in Power BI. Maybe you've licensed Fabric. Your team is being asked about Copilot, AI agents, or "what are we doing with GenAI" — and the honest answer is that nobody knows whether the underlying model would give good answers.
The sprint is a fixed, time-boxed way to find out without committing to a six-month programme first.
Seven to ten days.
Real artefacts.
- A ranked inventory of your existing semantic models, reports, and data sources, with governance red flags flagged.
- A shortlist of 5–10 candidate AI/decision use cases matched to specific business decisions, scored by impact and readiness.
- A priority-ordered remediation plan — what to fix in the model, the measures, and the governance before rolling out Copilot or an AI agent.
- A working Copilot demo against your highest-readiness model, so leadership can see what "good" looks like on their own data.
- An executive read-out deck and a technical handover document — both yours to share internally without re-engaging us.
- A written recommendation on whether to proceed to an Ask Your Data Workflow or Predictive Power BI engagement, with indicative scope and pricing.
Three phases.
Fixed timebox.
Structured interviews with finance, ops, or commercial leads. Tenant and workspace audit. Semantic model review using Claude Code + Fabric MCP, so we inventory thoroughly without eating your team's calendar.
Readiness scoring of each model and use case against governance, measure quality, and AI-fitness criteria. Draft of the remediation plan. A working Copilot prototype against the strongest model.
Technical walkthrough with your data team. Executive readout with your buyer. Written handover. Indicative scope for the next engagement, if you want one.
The questions buyers actually ask.
Why would we need this if we already have Power BI?
Having Power BI does not automatically mean your data is ready for AI workflows. Most mid-market models were built to render reports, not to answer questions. Copilot and agents amplify whatever is already in the model — including the gaps. The sprint tells you where those gaps actually are.
Is this a strategy deck or actual work?
Actual work. We inventory your models, score them, build a working Copilot demo, and hand over documentation. The deck at the end summarises what we built — it is not the deliverable.
What if the sprint tells us we're not ready?
That is a valid outcome and we will say so clearly. Knowing you need six months of model remediation before Copilot is worth rolling out is a better answer than finding out after you have already rolled it out. No pressure to proceed to a larger engagement.
Do you work inside our tenant or yours?
Yours. Claude Code and any MCP tooling runs against your Azure tenant with access you control. No data egresses to external APIs unless you explicitly authorise it.
How is the sprint priced within the $22K–$35K range?
Range depends on workspace size, number of use cases to evaluate, and whether you want the working Copilot demo included. We confirm final fixed fee in writing before we start.
Show us the decision
that still takes too long.
A free 45-minute call. Bring a workflow, a reporting pain, or a trust issue — we'll tell you quickly whether this is a real fit.
Request a scoped proposal