Back to Portfolio
MockRoom • AI Career Diagnostic & Engineering Loop Calibrator

System Design Arena & Resume Bullet Word-Diff Clinic

Benchmark your architecture trade-offs across a 5-axis Staff rubric, transform passive resume bullets into quantified XYZ impact statements, and hand off your diagnostic scorecard directly into a 1:1 mentorship ticket or live session.
Target Bar:
Select System Design Track
30:00
Prompt, Traffic Scale & Hard ConstraintsSenior Loop
Design the end-to-end checkout, stock reservation, and payment confirmation architecture for a flash drop. Detail capacity sizing, hot-key sharding or token-bucket reservations, idempotency keys across retries, transactional outbox/saga orchestration, and operational telemetry.
Traffic Profile45k RPS read burst, 6.5k RPS checkout attempts, 93:7 read-to-write ratio
Target SLOs & Invariantsp99 reservation < 140ms, 99.99% checkout availability, strict zero-oversell invariant
Non-Negotiable Failure & Scale Constraints:
1. Single hot SKU cannot serialize all 6,500 concurrent checkout transactions onto one PostgreSQL row lock.2. Payment gateway calls exhibit p99 tail latency up to 4.5s and intermittent 504 timeouts after capture.3. Abandoned cart holds must expire deterministically within 8 minutes without leaking reserved stock.
Architecture Building Blocks & Design Scratchpad
Toggle the topology components you are committing to (3/6 selected):
Written Architecture & Trade-Off Justification0 words
Evaluates 5 weighted axes + Staff follow-up probes
5-Axis Staff Rubric ScorecardStaff-Ready Bar
Composite Rubric Score vs Senior Bar
93/ 100 (Target: 72+)
Deterministic Staff Rubric
Your "Flash-Sale Inventory Hold & Idempotent Payment Saga" architecture clears the Senior bar (93/72) with strong signal in Consistency & Idempotency (100/100). Next, pressure-test Observability & Operations (83/100) under live failure-mode probing.
Capacity & SLOs96/100Data & Sharding86/100Idempotency100/100Backpressure100/100Observability83/100
Your Design Polygon
Target Bar (72/100)
Capacity & SLO Math96/100
Detected: Component: Edge Waiting Room & Token Bucket Limiter • Component: Virtual Sharded SKU Sub-Buckets (N=32)
Data Model & Partitioning86/100
Detected: Component: Virtual Sharded SKU Sub-Buckets (N=32) • Component: Client + PSP Idempotency Key Ledger
Consistency & Idempotency100/100
Detected: Component: Client + PSP Idempotency Key Ledger • Component: Transactional Outbox & Saga Compensator
Fault Tolerance & Backpressure100/100
Detected: Component: Edge Waiting Room & Token Bucket Limiter • Component: Transactional Outbox & Saga Compensator
Observability & Operations83/100
Detected: Component: W3C Traceparent + Stock Conservation Alerts • Distributed tracing (W3C traceparent / OpenTelemetry)
Strengths Detected
• Consistency & Idempotency (100/100): Component: Client + PSP Idempotency Key Ledger + Component: Transactional Outbox & Saga Compensator.• Fault Tolerance & Backpressure (100/100): Component: Edge Waiting Room & Token Bucket Limiter + Component: Transactional Outbox & Saga Compensator.• Capacity & SLO Math (96/100): Component: Edge Waiting Room & Token Bucket Limiter + Component: Virtual Sharded SKU Sub-Buckets (N=32).• Data Model & Partitioning (86/100): Component: Virtual Sharded SKU Sub-Buckets (N=32) + Component: Client + PSP Idempotency Key Ledger.• Observability & Operations (83/100): Component: W3C Traceparent + Stock Conservation Alerts + Distributed tracing (W3C traceparent / OpenTelemetry).
Critical Blind Spots to Address
• Observability & Operations (83/100): Missing Multi-burn-rate SLO alerts or meta-monitoring.
3 Targeted Staff-Interviewer Follow-Up Probes
Q1: Which exact continuous ledger invariant query and SLO burn-rate window trigger an automated canary halt during a live flash drop?Q2: When 98% of virtual SKU shards hit zero remaining units and the final 100 units sit on 2 shards, how do you rebalance or route without false out-of-stock rejections?Q3: Show how you size Redis/PostgreSQL IOPS and worker concurrency to sustain 6,500 checkout writes/sec while holding p99 reservation latency below 140ms.
1-Click Coaching & Architecture Handoff

Turn Your Diagnostic Gaps Into a Targeted 1:1 Session or Mentorship Thread

Matched directly against active portfolio services from `/api/public/tickets` based on your lowest rubric axes.
Cloud & DevOps~24h SLA
High-Availability Kubernetes, Containerization & OpenTelemetryDeepens Observability & Operations (83/100) — w3c traceparent, red/use metrics, canary rollback, chaos drills.
Databases & HA~24h SLA
Database Architecture, Query Tuning & High AvailabilityDeepens Data Model & Partitioning (86/100) — schema, sharding/indexing, hot-key mitigation, retention.
Tutoring & Mentoring~48h SLA
1-on-1 Tutoring: SAT/IGCSE Maths, Physics, CS & EnglishDeepens Capacity & SLO Math (96/100) — traffic math, read/write ratio, p99 latency & availability slos.