Read chargebacks-weekly-export.csv from my workspace. For each open case, create an evidence preparation package: (1) Match each case to its Visa/Mastercard reason code category (fraud, consumer dispute, processing error, authorisation). (2) For each case, note the Card_Scheme (Visa/Mastercard/AMEX) and calculate the response deadline using public scheme rules: Visa 30 days from notification, Mastercard 45 days, AMEX 20 days. (3) Draft a rebuttal outline based on the reason code — what evidence is typically needed (transaction receipt, delivery proof, communication log, AVS match). (4) Flag cases approaching THEIR scheme's deadline within 3 days as URGENT — do not apply a single deadline to all schemes. (5) Create an Excel workbook: Sheet 1 — All cases with scheme, deadline, rebuttal outlines and evidence requirements. Sheet 2 — URGENT cases with a step-by-step action plan. Sheet 3 — Summary: cases by category and scheme, total amount at risk, win probability estimate based on reason code. Add a narrative summary recommending which 5 cases to prioritise this week and why.
Read hubspot-deals-export.csv from my workspace. For each unique Current_Provider (competitor) in the deals list: (1) Create a 1-page battle card in a single Word document. Each card should include: competitor name, number of your deals competing against them, typical deal size range, industries where you're competing, and 5 key talking points — what your company does better, what the competitor's known weaknesses are (based on industry patterns for payment providers), common objections merchants raise when switching, and a suggested opening line for the call. (2) Add an executive summary ranking competitors by total deal value at stake. (3) Flag any deals that have been in the same stage for more than 30 days as STALE — these need a different approach.
Read zendesk-tickets-48h.csv from my workspace (50 resolved tickets). Analyse the tickets and: (1) Identify the top 10 most common question patterns — group tickets by similar subject/issue type. (2) For each pattern, draft a knowledge base article: title, summary, step-by-step resolution, common variations, escalation triggers (specific conditions that mean the agent should stop troubleshooting and escalate to a specialist team), and when to escalate. (3) Flag any questions where different tickets got contradictory resolutions — these need a team decision on the correct answer. (4) Create a Word document with all 10 draft articles, ready to paste into your wiki. (5) Add an appendix: ticket volume per category, percentage of tickets each article would deflect, estimated time saved per week if these articles existed.
I have two files: screening-alerts.csv (30 alerts) and merchant-profiles.csv (20 merchant profiles). Simulate a weekly alert triage: (1) Match each alert to its merchant profile by Merchant_ID. (2) Pre-categorise each alert as FALSE POSITIVE (name match only, no other risk indicators, low-risk industry), NEEDS REVIEW (some risk factors present, medium-risk industry or elevated volume), or ESCALATE (multiple risk factors, high-risk industry, PEP/sanctions match, or merchant CDD review overdue >12 months). (3) For each alert, draft a 2-3 line investigation rationale explaining WHY you assigned that category — this is the part that takes the most time to write manually. (4) For each ESCALATE alert, add a recommended next step (e.g., request source of wealth documentation, refer to MLRO, trigger enhanced due diligence). (5) Create an Excel workbook: Sheet 1 — All alerts categorised with rationale. Sheet 2 — ESCALATE cases with summaries and next steps (for team lead review). Sheet 3 — Statistics: false positive rate, alerts by category, merchants with multiple alerts. Add a narrative memo summarising this week's risk landscape.
I have two settlement files: bank-settlements.csv and internal-settlements.csv (40 rows each). Reconcile them: match by Merchant_ID + Settlement_Date + Reference_Number. Create an Excel workbook with: Sheet 1 - Matched transactions (amounts equal). Sheet 2 - Variances (amounts differ) with Difference column, sorted by largest variance first, RED for >$500, YELLOW for $100-500. Sheet 3 - Missing from bank (in internal but not bank). Sheet 4 - Missing from internal (in bank but not internal). Sheet 5 - Summary with formulas: total matched, total variances, total missing, net discrepancy, largest single variance.
Read process-description-raw.txt from my workspace. This contains notes about our merchant onboarding process across multiple stages. Create TWO outputs: (1) A process analysis Word document that: maps each stage of the onboarding journey, identifies handoff points between teams, flags steps that are manual but could be automated, pays special attention to the CDD/due diligence stage (in most FinTech companies this is where onboarding gets stuck), estimates where the biggest time sinks are based on the process description, and recommends 3 specific bottleneck fixes. (2) A "New Starter Guide to Onboarding" — a 2-page plain-English document that someone joining the ops team could read on day 1 to understand: what the process is, who does what, where things get stuck, and who to ask when something goes wrong. Write it in a friendly tone, not corporate-speak.
Pick the department challenge that best matches your role. If none fit exactly, use the Support (knowledge base builder) or Finance (settlement reconciliation) challenge — these are the most universal.
