CClaude Cert Prep
Retail · 13 linked questions

Peak-Season Support at a Global Retailer

A global e-commerce brand is rolling out a Claude-powered customer-support and product-recommendation system. Follow the team from launch through a 10x Black Friday surge, a cost-mix review, and a brand-safety incident — making the architecture, model-routing, and guardrail decisions that keep it fast, cheap, and safe.

One evolving scenario Questions build on the context Rate your confidence
How to use it · Read the brief, then work top to bottom — the situation evolves as you go. Rate your confidence before submitting; on review, the “ confident & wrong” questions are the misconceptions worth fixing first. A study tool, not the graded exam.

The scenario

NorthMarket, a retailer operating in 14 countries, is replacing a rules-based chatbot with a Claude system that handles real-time shopper chat and bulk back-office copy generation. The launch team must right-size models, control cost and latency at extreme peak, and enforce brand-safety guardrails — all before Black Friday.

Architecture
A 'router' agent hands off to specialist sub-agents (returns, order-status, product-reco, copywriting); the team assumed multi-agent was mandatory. Real-time chat is synchronous; catalog/marketing copy is bulk back-office.
Existing stack
Next.js storefront, existing rules chatbot, product catalog in Postgres, message queue for back-office jobs, Datadog for observability. No LLM in production yet.
Scale
~40K support chats/day at baseline; Black Friday peaks historically hit 10x overnight. Back-office: ~2M product descriptions and marketing snippets regenerated seasonally.
Regions
Shoppers in North America, EU, and APAC; EU traffic carries data-residency expectations. Traffic follows the sun with sharp regional spikes.
Budget
Cost-sensitive: finance caps the pilot at a fixed monthly spend and wants per-conversation cost tracked. Overruns trigger a mandatory model-mix review.
SLA
Real-time chat: first-token < 1s, full response typically < 5s. Bulk copy: no per-item latency target, must finish within a 24-hour regeneration window.

Keep this context in mind — later questions build on it, and the situation evolves as you go.

0 of 13 answered
  1. P1 · Q1Model right-sizing across two workloads
    The team wants a default model-routing policy for launch. Real-time chat needs sub-second first-token latency and tight cost control; bulk copy generation is quality-sensitive but not latency-sensitive. What is the best default policy?
    How confident are you?
  2. P2 · Q2Cost levers that respect latencySelect 2
    Finance wants launch cost down without hurting the real-time chat experience. Which TWO choices reduce spend while preserving chat latency?
    How confident are you?
  3. P4 · Q3Latency for interactive chat
    Product wants the chat to feel instant, targeting first-token under 1 second. Which delivery approach best meets that for the synchronous chat path?
    How confident are you?
  4. P7 · Q4Guardrail failure mode
    Before launch, the team designs a brand-safety guardrail that screens generated copy for prohibited claims and off-brand tone. If the guardrail service is unavailable or errors, what should happen?
    How confident are you?
  5. The situation changes

    Black Friday: traffic 10x overnight

    Peak weekend arrives and support chat volume jumps ~10x versus baseline within hours, concentrated in EU morning and US evening windows. Per-conversation cost and p95 latency both start climbing, and finance is watching the daily spend line in real time. The team must hold the chat SLA without breaking the budget.

    P1 · Q5Preserving SLA under surge
    At 10x volume, chat p95 latency and cost per conversation are both rising. What is the best first move to hold the SLA within budget?
    How confident are you?
  6. P2 · Q6Bulk work during peak
    During the peak, marketing asks to regenerate 300K promotional snippets 'as soon as possible.' There is no per-item latency need and results are wanted within the day. How should this run?
    How confident are you?
  7. P4 · Q7Latency and cost knobs at peakSelect 2
    The team has minutes to shave chat cost and latency during the surge without a redeploy. Which TWO adjustments help most?
    How confident are you?
  8. P3 · Q8Multi-region routing
    EU traffic peaks first and carries data-residency expectations, while US and APAC peak later. How should the team handle regional routing during the surge?
    How confident are you?
  9. The situation changes

    A cost spike forces a model-mix review

    Black Friday is over, but the finance dashboard shows per-conversation cost ran far above plan — enough to trigger the mandated model-mix review. Digging in, the team finds two things: a large share of chats were served by Opus, and the prompt-cache hit rate was near zero for much of the weekend. The router's design is now under scrutiny.

    P5 · Q9Diagnosing the cost overrun
    The review needs to find the root cause of the overrun before changing anything. Which investigation is most likely to explain the spike?
    How confident are you?
  10. P1 · Q10Architecture reassessment
    The review also finds the router-plus-specialist multi-agent design added several model hops per conversation, inflating tokens and latency, and that most conversations only ever needed one specialist. What should the team do?
    How confident are you?
  11. P2 · Q11Structural cost reductionsSelect 2
    Coming out of the review, the team wants durable cost reductions for the next peak. Which TWO changes give the most structural savings?
    How confident are you?
  12. The situation changes

    A brand-safety incident in generated copy

    A week later, a customer screenshots a promotional snippet that made an unverifiable health claim about a product — one of the 300K batch-generated snippets that went live. Legal and brand are alarmed. Investigation shows the batch copy pipeline published straight to the storefront without passing the brand-safety guardrail that the interactive chat path uses.

    P7 · Q12Root cause of the incident
    What is the root cause of the brand-safety incident, and the correct fix?
    How confident are you?
  13. P6 · Q13Hardening the pipeline post-incident
    Beyond adding the screening, what is the best next action to prevent a repeat and catch regressions early?
    How confident are you?