System Design Interview Questions

Darshak S. - "Clarifying Questions: There are several different types of therapists, do you have a particular therapist in mind? Like Mental health, physical health, food and nutrition etc? Reply: You can decide And do we know what country are we building this product, since that will help me narrow down the scope of the product? Reply: USA Approach: Define the product vision and the goal that we want to achieve Target Audience is for this particular product What problems do these user currently"See full answer

Brody C. - "Clarifying Questions: When you say Language Learning app, you're referring to a user wanting to learn a new language through a Meta property - not NLP tech correct? -- Yes Is this a stand-alone app, or part of Facebook or another Meta app? -- You choose Are there any restrictions like budget, or time for this learning app? -- Nope. Give me a moment to think about how I want to structure this answer: Mission // Goal of Language Learning App Competition Audience Segments Pain Points "See full answer

Anonymous - "Functional requirements: (a) location tracking (b) check-in at nearby place (c) view historical checkins with how much time user stayed there (d) Mark checkout when out of proximity (e) analytics - checkins by user at given loc, top n places Out of scope: (a) checkin sharing with followers Non functional: (a) low latency (b) high availability (c) eventual consistency Scale: (a) QPS: 10M DAU * 2 checkins /86400 seconds (b) Yearly Data Volume: 10M* 0.1 KB *365 days HLD: user -> location servic"See full answer

Jaya S. - "Understood the problem statement by confirming details with interviewer. Approached by listing down all the basic features being provided by the platform. Platform -> Organizations can register themselves. One org can have more than one recruiters. Recruiters create jobs on the org portal. Candidate can apply to the jobs. Recruiters can accept and reject the applications. Categorized 4 parent tables by taking hints in between. The tables were USER -> ID, Name, Phone No, mail ID, Profile Des.("See full answer

Jonas P. - "Searching for podcasts over topics is quite easy nowadays. However, usually we either don't find any podcasts interesting enough or are overwhelmed with choice. What really indicates what podcast we want to listen to at the moment is our mood, energy level etc. Additionally, we typically have some requirements, like being on the train and having one hour to spend, or only a five-minute interval while preparing breakfast. As such, I want an interface that lets me search for length. I want my"See full answer

Zain R. - "I've worked on projects not quite like this, but very similar, in the past - I'll borrow from that to answer this: The Broader Context this problem doesn't specify the type of data we're working with, or how it's being ingested to align with my personal background, I'll assume a picture that lends this problem well to being a computer vision (abbreviated "CV") related question: let's say we have a conveyor belt in a waste facility, which sequentially carries a stream of waste w"See full answer

Chaitali M. - "Clarifications: Is it the voice only Alexa devices or includes echo show (video)? Goal: To have people to safety quickest Users: Alexa owner At the house when somebody breaks in Family Single adult/seniors Outside the house at the time of break in Police Department Pain points/user group: Owner in the house at the time of burglary Inform all the house members at the earliest. Send SOS out on all/selective Alexa devices to ask family members to reach a safe place 2"See full answer

Rasika D. - "Message queue is like a buffer. Its has 3 parts: Producer -> which creates a message and sends it to the queue. Queue -> which acta like a buffer pipe and stores the message. Subscriber -> which processes and releases messages one by one in FIFO manner."See full answer

Jiale L. - "I’d open by confirming we’re designing object storage for arbitrary binary objects—images, backups, data lake files—not block volumes for VMs unless you redirect me. Actors: developers and services inside Amazon’s ecosystem, multi-tenant, millions of buckets, trillions of objects. MVP API: create bucket, PUT object, GET object, DELETE, list objects with prefix, HEAD for metadata. I’d ask about versioning, encryption at rest default, and whether"See full answer

Anonymous Hornet - "To handle the non-uniform sampling, I'd first clean and divide the dataset into chunks of n second interval 'uniform' trajectory data(e.g. 5s or 10s trajectories). This gives us a cleaner trajectory data chunks, T, of format (ship_ID, x, y, z, timestamp) to be formed. For the system itself, I'd use a generative model, e.g. Variational AutoEncoder (VAE), and train the model's 'encoder' to produce a latent-space representation of input features (x,y,z,timestamp) from T, and it's 'decoder' to pred"See full answer

Yulia D. - "What age of kid we are speaking about? Lets assume 5-10 years what messenger app are we talking about? you can choose Who is creating the feature? you can choose Finally, do we have any constraints? no, please, just imagine ok. so I think company's goal should be giving the tools for building communities among the youngest ones. First I want to issue what is Kids messenger app. I have kid of 5 years old and among my friends there are only few with gudgets that can provide messenger f"See full answer

Amin S. - "We can implement Mall psychology, use the monkey effect, FOMO, top-rated stuff from the same restaurants, bundle offers like couple or group offers, and prompt popups for free delivery if you complete X$ amount by adding X item. We can lure customers by giving free subscriptions to Spotify, Netflix, etc. by collaborating with these giants. This is a long-story-short overview of how we can increase the average basket size. Tracking telemetry and refining can make a difference obviously..."See full answer