Northslope Case Coach

What's your case?

Paste in the case prompt — anything from a one-liner to a full scenario. The coach will guide you through each phase of the decomposition framework.

Case Prompt

Paste the full case scenario below

Available Data Columns

Click to tag the data fields mentioned in the case (or type them)

timestamp user_id order_value product_id device_type region payment_method customer_id arrival_time clinician_id room_id fare_amount driver_id distance picker_id dispatch_time sku_count account_tier logins_30d usage_count turbine_id output_kwh wind_speed temperature maintenance_flag operator_id approval_decision default_status loan_amount incident_type location_zone unit_base_location machine_id material_batch_id temperature_log qc_result payer_type diagnosis_code net_revenue_per_visit seniority_level features_accessed session_duration

Case Type

What kind of problem is this?

AI will guide each phase

Phase 1 — Reframe the Brief

🔑 Before touching any data, challenge what the prompt is actually asking. Strong candidates spend 60–90 seconds here.

What hidden assumptions are baked into this brief?

What clarifying questions would you ask the client first?

Phase 2 — Decompose the Problem

🔑 Break the top-level metric or problem into its component parts. Never diagnose the aggregate — find which component is failing.

What are the key stages or components of this problem?

Which segment dimensions should you cut across?

By time period

By user segment

By geography

By product / SKU

By channel

By device type

By individual (person)

By order complexity

By account tier

Phase 3 — Build Hypotheses

🔑 For each component you've identified, generate a ranked list of falsifiable hypotheses. Each must be testable with the data you have.

Your hypotheses (ranked — most likely first)

What type of failure is this? (select all that apply)

Demand increase Supply / capacity drop Discovery failure Value failure Segment fit mismatch Process / design fault External event

Phase 4 — Data Strategy

🔑 For each hypothesis, name the specific metric or query that confirms or kills it. Don't just collect data — decide what each data point proves.

What metrics or queries would you derive from the available data?

What data gaps exist? What are you assuming about data quality?

Phase 5 — Check-in Moment

🔑 Check in with the interviewer 2–3 times per case. This is moment 2 (after hypotheses). Show confidence, invite course-correction, protect from wrong paths.

Write your check-in statement to the interviewer

What new information would change your current hypothesis ranking?

Phase 6 — Recommendation

🔑 End with a decision, not just analysis. Tier your recommendations by speed of implementation. Name what you'd monitor going forward.

Immediate Actions (this week)

Short-term Fixes (2–4 weeks)

Ongoing monitoring — what metrics would you track?

Your final recommendation in one sentence

⧗

Complete Phase 6 to generate your full debrief and scores.