Benchmarks

This regex is supposed to match ISO dates but also matches '2024-13-45'. Fix it: /\d{4}-\d{2}-\d{2}/

Write a React useDebouncedValue hook in TypeScript. It should accept a value and a delay in ms, update only after the value stops changing, clean up its timer on unmount, and be fully typed with generics.

Refactor this Node.js module from nested callbacks to async/await, preserving error handling semantics and adding proper TypeScript types:

```js
function loadUser(id, cb) {
  db.query('SELECT * FROM users WHERE id = ?', [id], (err, rows) =…

Design a distributed rate limiter for an API gateway serving 50k requests/sec across 12 edge regions. Compare token bucket vs sliding window at this scale, explain where state should live, how you handle clock skew, and what happens during …

I'm building a real-time gaming leaderboard with 2M daily active users, frequent score updates, and global top-100 reads. Should I use Postgres, Redis sorted sets, or something else? One paragraph on why.

We have a Rails monolith processing 30k orders/day. I need a phased migration plan to event-driven microservices: what to carve out first, how to keep the monolith authoritative during transition, how to avoid dual-write inconsistencies, an…

Extract the name, title, company, phone, and email from this signature as JSON:

Best,
Priya Ramanathan
Senior Director, Platform Engineering
Hexawave Systems
+1 (415) 555-0173
priya.r@hexawave.io

From this invoice text, extract each line item with description, quantity, unit price, and total as a JSON array:

INVOICE #4482
Web hosting (annual) — 1 × $240.00 — $240.00
SSL certificate — 2 × $45.00 — $90.00
Overage bandwidth 40GB — 1 ×…

Convert every date in this text to ISO 8601 and list them: 'The lease starts March 3rd, 2025, with a review on 9/15/2025 and termination no later than 2026-02-28.'

Summarize this support ticket in one sentence for a triage dashboard: 'Hi, since the update yesterday around 2pm our team of 14 people cannot log in via SSO, we get redirected in a loop. Regular password login works. This is blocking all wo…

Summarize the following operations report into 5 bullet points for an executive readout:

Quarterly operations review — Northbridge Logistics. Revenue grew 8% quarter over quarter, driven by same-day delivery expansion in three metro areas.…

Turn these meeting notes into action items with owners:

'Ana: launch moved to the 14th because payments aren't ready. Boris will finish Stripe webhook retries by Friday. Priya to confirm the email copy with legal — they flagged the discoun…

Translate these UI strings to Spanish, keeping placeholders untouched: 'Save changes', 'Discard draft', '{count} items selected', 'Your trial ends in {days} days'.

Translate to natural, polite Japanese suitable for an end-user error dialog: 'We couldn't process your payment. Your card was declined. Please try a different payment method or contact your bank.'

Translate this product blurb to German, preserving the confident tone and adapting idioms naturally rather than literally: 'Stop babysitting your deploys. Ship on Friday at 5pm and sleep like a baby — our rollback button has your back.'

A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost? Answer with the amount only.

We can either ship SSO this quarter (requested by 4 enterprise prospects worth ~$220k ARR combined) or rebuild onboarding (current activation is 31%, industry median ~45%). Engineering capacity: one team, one quarter. Which do we pick and w…

Five engineers (Ada, Babbage, Church, Dijkstra, Hopper) each own exactly one of five deploy slots (Mon–Fri) and one language (Go, Rust, Python, TS, Haskell). Clues: Ada deploys before Church. The Rust user deploys on Wednesday. Dijkstra use…

Three people split a $86.40 dinner with 18% tip. Two had equal shares, the third paid exactly double. What did each pay? Round to cents.

Our MRR is $48k growing 7% month over month. Churn is 3.5% of MRR monthly, applied to the prior month's base. Model the next 12 months month-by-month and tell me the month we cross $100k MRR. Show the formula you used.

A monitoring pipeline has three independent stages with false-negative rates 2%, 5%, and 1%. An incident must pass all three checks undetected to be fully missed. (a) Prove the probability of a fully missed incident. (b) We can halve exactl…

Read the following operations report carefully. Somewhere in it, two statements contradict each other about the Reno facility. Quote both and explain the contradiction:

Quarterly operations review — Northbridge Logistics. Revenue grew 8% q…

Write a 150-word executive brief from this report, focusing only on financial risks and the two most urgent operational problems:

Quarterly operations review — Northbridge Logistics. Revenue grew 8% quarter over quarter, driven by same-day…

Analyze this report end to end. Identify every claim about cost savings, cross-reference each against the stated risks and overruns, and produce a net assessment of whether the reported savings are credible. Structure the answer as: claims …

What's the weather in Lisbon right now, and should I bring a jacket tonight? Use the weather tool.

Look up the customer with email deniz@acme.co, find their last 3 invoices, check if any are overdue, and if so draft a polite payment reminder email referencing the specific invoice numbers. Use the available tools in the right order.

For our top 10 accounts by MRR: pull current usage metrics, compare against their plan limits, flag any within 15% of a limit, then for each flagged account draft an upsell note customized to their usage pattern and file it as a CRM task as…

Write a 3-line Slack update for #eng: the search reindex finished overnight, latency is back under 200ms, and we're monitoring for 24h before closing the incident.

Write a changelog entry for: new CSV import with column mapping, 3x faster dashboard loads, and a fix for the timezone bug in scheduled reports. Tone: clear, a little confident, no hype words. Include a 'Why it matters' line per item.

Write a 900-word launch post for our new usage-based pricing. Cover: why we switched (customers hated seat math), how the migration works (grandfathering for 12 months), a worked cost example for a 20-person team, and an FAQ. Voice: direct,…