Interrogate the estate
Use SQL to establish grain, point-in-time availability, and reconciled evidence before reaching for a model.
ANALYTICAL WORK, UNDER OPERATING CONDITIONS
The Analyst places you inside Meridian Living Systems, a fictional products-and-services company. Across nine full assignments and a permanent weekly rotation of 16 Priority Briefs, you step into different analyst roles and confront what clean tutorials remove: unclear requests, competing clocks, unreliable grains, operational limits, and accountable handoffs.
WORKSPACE NOTE / Complete assignments in a current laptop or desktop browser.
YOUR EMPLOYER, FOR THE DURATION OF THE WORK
The Analyst is the training environment. Meridian Living Systems is the fictional company inside it: a multi-region operator selling connected-home products and subscriptions, dispatching field service, running warehouses and commerce, and supporting customers after the sale.
Meridian has grown through acquisitions, platform migrations, new models, and operational change. Its data reflects that history. Across nine assignments, you move between teams and moments in the same company. The systems, customers, devices, branches, definitions, and consequences belong to one connected estate—not nine unrelated classroom datasets.
AI assistance is optional. The workbench remains fully usable without it, and generated advice is never treated as evidence.
Across every assignment and Priority Brief, a small local model can explain a visible error, compare a highlighted recommendation with the request, or pressure-test a highlighted technical claim. It is not an open chat, grader, or answer key—and no account is required.
Every assignment and Priority Brief can produce a structured Markdown packet for an external AI tool. You choose the question and included material. The packet is created locally and leaves the workbench only when you deliberately share it.
Use AI to explain, question, and surface checks—not to certify a result or decide what the evidence means. Run the analysis, inspect the sources, verify important guidance, and make a handoff you can defend.
The sequence moves from metric reconciliation to production-grade model risk. Complexity rises, but judgment begins in assignment one.
Reconcile two conflicting satisfaction figures before the executive review.
Foundation02Certify Q2 orders, revenue, and fulfillment timing after an acquisition cutover.
Investigation03Determine whether a mobile-navigation experiment warrants rollout.
Decision04Recommend global rollback, scoped containment, or monitored continuation during a storm.
DecisionDATA RESPONSIBILITY / INCOMING
Customer text may improve routing, but the corpus contains later messages, internal notes and synthetic PII. Define the useful data boundary without exporting the risk.
The self-guided manual supplies readiness gates, honest workload routes, a repeatable assignment cadence, stuck protocols, and a spoiler-controlled way to review your own work.
The Career Readiness Evidence Guide connects assignments to carefully bounded NACE competency practice, portfolio review, truthful résumé and career-profile language, interview preparation, and ready-to-use career-center formats.
Planning guidance, workload bands, assessment boundaries, and local review tools sit in a deliberately separate instructor desk so the learner route stays focused on the work.