Design a Rate Limiter
MEDIUMDesign and implement a rate limiter that can handle millions of requests. Support multiple algorithms: token bucket, sliding window, fixed window. Consider distributed systems.
Practice your Google interview before the real one.
Apex conducts realistic mock interviews, mimicking real loops from Google, Meta, Amazon, and more, then scores you with the rubrics their interviewers use.
- Used by 1000+ engineers everyday
- Scored against each company's hiring bar
- Code editor and whiteboard, like the real loop
150+ companies
- Amazon
- Uber
- Apple
- OpenAI
- Anthropic
- xAI
- Meta
- Microsoft
- Waymo
- Perplexity
- Goldman Sachs
- Netflix
- C3.ai
- Abnormal Security
- IBM
- Oracle
- Citadel
- Square
- Doordash
- Dropbox
- Jump Trading
- JP Morgan
- Capital One
- Bloomberg
- ByteDance
- Adobe
- TikTok
- Visa
- MongoDB
- Snapchat
- Two Sigma
- Verkada
- Grammarly
- TradeDesk
- AFFIRM
- Anduril
- ARM
- Nextdoor
- Rippling
- Glean
- PayPal
- Airbnb
- Intuit
- Palantir
- Agoda
- Anyscale
- Applied Intuition
- Axon
- BaseTen
- Benchling
- Canva
- Cisco
- Cohesity
- Confluent
- CoreWeave
- CrowdStrike
- Databricks
- Datadog
- DE Shaw
- Disney
- Ericsson
- Expedia
- Figma
- Flexport
- Geico
- Gusto
- Headlands
- Instacart
- Jane Street
- Lead Bank
- Lyft
- Millennium
- Moveworks
- Okta
- Optiver
- Oscar Health
- Qualcomm
- Ramp
- Replit
- Roblox
- ServiceNow
- Sigma Computing
- Snowflake
- SoFi
- Squarespace
- Temporal
- Twilio
- Valon
- Vanta
- Zipline
- Airtable
- Atlassian
- Bobyard
- CodeSignal
- Cursor
- eBay
- Headway
- HubSpot
- Notion
- Nuro
- Nvidia
- Otter.ai
- Patreon
- Plaid
- Point72
- Retool
- Rubrik
- Salesforce
- Samsara
- Toast
- Whatnot
- Block
- Brex
- Docker
- Faire
- Hugging Face
- Robinhood
- Sig
- Spotify
- Tesla
- Tower Research Capital
- Veeva
- Zip
- Chime
- Discord
- Mastercard
- Stripe
- Zoom
- ZipRecruiter
- Coinbase
- Scale AI
- Zoox
- Cohere
- Clay
- Voleon
- Attentive
- Cartesia
- CloudKitchens
- Coursera
- Decagon
- Harvey
- Inworld
- Quizlet
- Braze
- Navan
- Parafin
- Wayfair
- Yahoo
- Circle
- Persona
- Sierra
- Stubhub
- Asana
Where engineers landed after practicing here
Senior SWE
$340k → $465k
Better feedback than my $200/hr interview coach. The AI caught mistakes I didn't even know I was making.
Junior SWE
$135k → $190k
Finally understood what interviewers actually want. After 20 mock interviews, I went from nervous to confident.
Senior SWE
$420k → $730k
Follow-up questions were brutal, exactly what I needed.
Senior SWE
$350k → $700k
30 mock interviews here prepared me for the real thing. Landed Netflix with a 100% TC increase.
Senior SWE
$390k → $485k
Went from failing phone screens to landing Anthropic.
Junior SWE
$145k → $205k
Passed Google L4 on my third attempt.
10,000+ mock interviews completed on Apex
You know how to code. Can you interview?
Most engineers who fail top-tech interviews could solve the problem. They lost points explaining it, defending it, or adapting when the interviewer changed the constraints.
LeetCode gives you problems. Apex gives you the interviewer.
How a session works
- 01
Pick your company
Questions, format, and scoring match that company's interview loop.
- 02
Interview out loud
Explain your approach, code in the editor or sketch on the whiteboard, and handle follow-ups as the interviewer pushes back.
- 03
See what would have cost you
Every note points to a moment in your transcript, like "You didn't state the O(n) extra space your hash map uses."
- 04
Get retested until it sticks
The gaps from this session come up again in your next interviews, until your scores show they're closed.
- GoogleInterview questions112 questions →
- MetaInterview questions71 questions →
- AppleInterview questions95 questions →
- MicrosoftInterview questions70 questions →
- NetflixInterview questions60 questions →
- TikTokInterview questions38 questions →
- UberInterview questions99 questions →
- AmazonInterview questions118 questions →
- OpenAIInterview questions95 questions →
- AnthropicInterview questions79 questions →
- PerplexityInterview questions68 questions →
- xAIInterview questions72 questions →
- OracleInterview questions51 questions →
Medium
Two Sum
Pythondef twoSum(nums, target): seen = {} for i, num in enumerate(nums): diff = target - num if diff in seen: return [seen[diff], i] seen[num] = i return []What's the time complexity of your solution? How would it handle duplicate values?
Coding Interview · Example
3.1/5
Borderline
- Correctness3.8
- Complexity2.4
- Code Quality3.3
- Communication3.0
- Problem Solving3.1
ChatGPT can quiz you. Apex interviews you.
- GoogleHard
Stream Window Extraction
- MetaHard
Score Tracker Analysis
- AmazonHard
Top-K Variant
- OpenAIEasy
IP Iterator
- AnthropicHard
Computer Use Security
Google
- Correctness
- Complexity
- Code Quality
- Communication
- Problem Solving
Meta
- Correctness
- Complexity
- Code Quality
- Communication
- Problem Solving
24/7
- system design
- ml
- behavioral/culture round
- design
- graph
- hash map
- intern
- string
- sde2
- tree
- data structure design
- dfs
- dp
- video interview
- bfs
- coding round 1 2
- debugging
- grid
- heap
- ml coding
- ood
- prefix sum
- reliability
- scheduling
How would this handle 10 million concurrent requests?
What's the trade-off?
- 04:43You
Should I consider distributed rate limiting?
- 05:02Interviewer
How would this handle 10 million concurrent requests?
- 05:21Interviewer
What's the trade-off?
See which skill is holding you back.
Every interview is scored on what your target company grades: correctness, complexity, communication, and more. Watch each score climb, and know exactly what to practice next. Up to 9 scored dimensions per round, tracked across 50+ topics.
Coding Interview
3.8/5
Strong Pass
- Correctness4.1
- Complexity3.6
- Code Quality3.8
- Communication3.9
- Problem Solving3.6
You work like you will on the day.
Write code in a real editor and sketch your system design on a whiteboard while you talk it through. The interviewer sees both and asks about them. No pasting code into a chat box.
def two_sum(nums, target): seen = {} for i, n in enumerate(nums): if target - n in seen: return [seen[target - n], i] seen[n] = iTest Cases
- Input
[2, 7, 11, 15], target = 9Output[0, 1] - Input
[3, 2, 4], target = 6Output[1, 2]
Every round, one place.
- 01
Coding
A real editor, starter code in 6 languages, and test cases. Code Review flags gaps in your solution before your interviewer does.
- 02
System design
Sketch your architecture on the whiteboard while you defend scale, bottlenecks, and trade-offs.
- 03
Behavioral
Rehearse leadership, teamwork, and conflict stories, scored on what your target company values.
- 04
Company-specific formats
Tailored to Google, Meta, Amazon, OpenAI, and more.
- 05
Full loops
Phone screens, virtual onsites, and pair programming, back to back.
What you get vs. the alternatives
Compare with
| Feature | LeetCode-style platforms | Peer mocks | Human coach | Apex Interviewer |
|---|---|---|---|---|
| Interviewer pushes back with follow-ups | No | Depends on the peer | Yes | Every session |
| Scored against your target company's bar | No | Rarely | Sometimes | Every session |
| Retests your weak spots | No | No | If you rebook | Automatically |
| Available whenever you are | Yes | Needs scheduling | Needs scheduling | 24/ |
| Cost | Free | Free | $200+/ | From $83/ |
Unlimited interviews for less than one coaching session.
Best value
$83 /month, billed $250 every 3 months
Save $50. Most job searches take longer than a month.
- Unlimited mock interviews
- All 150+ companies
- Coding, system design, and behavioral
- Code editor and whiteboard
- Feedback tied to your transcript
- Weak spots brought back until they're fixed
Questions, answered
How is Apex different from ChatGPT or Claude?
ChatGPT and Claude are great general tutors. Apex is built for one job: running your interview. It remembers how you performed and brings your weak spots back in later interviews, gives you a code editor and whiteboard the interviewer can see, and scores you against what your target company prioritizes.
Is there a free trial?
No. There's no free trial. Creating an account and taking the free coding diagnostic are free; unlimited mock interviews need a subscription. Every paid plan is unlimited, and the 3-month plan works out to $83/month, less than half the cost of a single session with a human coach.
What happens after I sign up?
Choose a plan, pick your target company and round, and start your first interview right away. No scheduling.
How is this different from LeetCode or HackerRank?
Those platforms check whether your code passes. Apex runs the interview around it: follow-up questions, feedback on how you communicate, company-specific scoring, plus system design and behavioral rounds. You find out how you sound, not just whether your code works.
Where do your interview questions come from?
From engineers who've interviewed at these companies, shared in interview reports, public forums (Reddit, Glassdoor), and by our community.
Can the AI really evaluate soft skills?
Yes. The interviewer listens to how you explain your approach, whether you discuss trade-offs, and how you handle follow-ups. Every note points to a moment in your transcript, like "You started coding before confirming the input constraints."
Does this work for senior and staff interviews?
Yes. Coding, system design, and behavioral rounds are calibrated from entry level to staff. At senior and staff levels, expect deeper probing on trade-offs, scale, and technical leadership.
Which companies do you support?
150+ companies, including Google, Meta, Amazon, Apple, Microsoft, Netflix, OpenAI, Anthropic, Uber, TikTok, Oracle, Perplexity, and xAI. We add more as our community reports new interviews.
Can I cancel?
Renews automatically. Cancel anytime before your next billing date. You can cancel from your account settings.
Do I need to schedule interviews?
No. Interviews are available 24/7.
The difference between failing and passing your tech interview isn't talent, it's preparation. Give yourself the unfair advantage.
Practice with AI interviews from top tech companies.