How would you design Mint.com (personal finance aggregator)?

Step 1: Use Cases, Constraints, Assumptions, Back-of-Envelope

Use cases

We'll scope to handle only the core flows that expose the bottlenecks; everything else is deferred.

• User links bank — OAuth/Finicity/Plaid flow, store encrypted tokens, test connection.
• Service syncs transactions — scheduled fetch (hourly/daily), normalize categories, dedupe by fit_id.
• User reads dashboard — aggregated balances, trends, budgets, search over normalized ledger.
• Service alerts — async rules (low balance, large spend) via notification queue.

Out of scope

• Bill-pay execution, tax filing, credit-score modeling, PII beyond financial data.

State assumptions

• 20M users, avg 3 linked institutions, 500 transactions/user/mo → 10B transactions total.
• Sync QPS: 20M*3/86k daily if hourly ≈700 fetches/s avg, bursty at top of hour 5k.
• Transaction 500 bytes → 10B*500≈5