weeeBox/mobile-system-design

A Simple Framework For Mobile System Design Interviews (iOS & Android)

This guide provides a framework for approaching mobile system design interviews, specifically for iOS and Android roles. The core idea is that interviewers focus on assessing your thought process and communication skills rather than expecting a perfect, production-ready solution within a limited time. By showcasing your strengths, you enhance your overall evaluation as a candidate.

Join the "Mobile System Design" Discord server for general discussions and feedback.

Disclaimer

This framework is inspired by "Scalable Backend Design" resources. Understanding this guide doesn't guarantee success. Interview structure varies based on the interviewer's style. Active dialog is vital – ask clarifying questions and avoid assumptions.

This guide does not reflect the interviewing policies from Google (or any other company).

Interview Process (45–60 min)

• 2–5 min: Introductions
• 5 min: Task definition and requirement gathering
• 10 min: High-level discussion
• 20-30 min: Detailed discussion
• 5 min: Questions for the interviewer

Introductions

Briefly share your relevant experience. Example: "Hi, I'm Alex. I've been developing for iOS/Android since 2010, focusing on frameworks and libraries. For the last 2.5 years, I led a team on the XYZ project, covering system design, planning, and technical mentorship." The purpose is to break the ice and highlight relevant skills.

Defining The Task

The interviewer presents the task, like "Design Twitter Feed." Determine the task's scope:

• Client-side Only: Design the app, given a backend and API.
• Client-side + API: Design both the client app and API. (Most Common).
• Client-side + API + Backend: Design the app, API, and backend. (Less Likely). Be transparent if your backend experience is limited. Focus on your mobile expertise and mention your high-level understanding of backend concepts from books, videos, etc.

Gathering Requirements

Categorize requirements into Functional, Non-Functional, and Out-of-Scope. Using "Design Twitter Feed" as an example:

Functional Requirements

Focus on high-value features (3-5).

• Users can scroll through an infinite feed of tweets.
• Users can "like" a tweet.
• Users can view comments for a specific tweet (read-only).

Non-Functional Requirements

These are critical for the product's overall success.

• Offline support: Users should be able to view cached tweets and interact with them even without internet connectivity.
• Real-time notifications: Notify users of new tweets, likes, and comments.
• Optimal resource usage: Minimize bandwidth, CPU, and battery consumption. Consider network optimization techniques and efficient data handling.

Out of Scope

Exclude features important for a real project but not in the interview scope.

• Login/Authentication
• Sending tweets
• Following/Retweeting
• Analytics

Providing the "Signal"

The interviewer evaluates your thought process. Show them:

• Your assumptions and how you stated them clearly.
• The rationale behind your feature choices.
• Relevant clarifying questions.
• Awareness of trade-offs and potential concerns.

Ask many questions and cover diverse aspects. Use a Breadth-First Search (BFS) approach, then ask the interviewer which area they want to explore in-depth. If they offer a choice, select a topic you know well.

Clarifying Questions

Ask these questions during the task clarification step:

• Do we need to support emerging markets?
  • Consider app size optimization due to low-end devices and expensive data. Prioritize caching and minimize network requests. On Android, look into using app bundles, on iOS investigate app thinning.
• What number of users do we expect?
  • High user count impacts backend load. Discuss Exponential Backoff and API Rate Limiting to prevent unintended DDoS.
• How big is the engineering team?
  • Smaller teams (2-4) require different considerations than larger teams (20-100). Focus on modularity and project structure to enable parallel development. Think about using a monorepo versus a multi-repo strategy.
• What are the target OS versions and device types?
  • This will drive decisions about architectural patterns (e.g. Compose/SwiftUI vs. older view-based systems), available APIs (e.g. for background tasks, data persistence), and necessary compatibility layers.
• Are there any specific performance benchmarks we need to achieve?
  • Understand acceptable latency for feed updates, image loading, and other key interactions. This will influence caching strategies, network optimization, and data serialization techniques.
• What level of data consistency is required?
  • Can we tolerate eventual consistency for likes and comments, or do we need strong consistency? This impacts decisions about offline behavior and data synchronization strategies.
• How frequently is the data updated?
  • If tweets are very frequent, then a pull-based refresh strategy might overwhelm the server. Push notifications or SSE might be more appropriate.
• What are the expected types of media?
  • Image sizes and formats will influence caching and networking strategies. Consider supporting modern formats such as webp or AVIF. Video will require different considerations, such as streaming protocols (HLS, DASH), transcoding, and CDN support.

High-Level Diagram

If the interviewer is satisfied, ask if they want to see a high-level diagram.

Server-side Components:

• Backend: The overall server infrastructure. Focus on client-side concerns.
• Push Provider: Infrastructure for mobile push notifications. Delivers payloads from the backend.
• CDN (Content Delivery Network): Delivers static content (images, videos) to clients efficiently.

Client-side Components:

• API Service: Abstract client-server communication. Provides a clean interface for interacting with the backend. Consider using a library like Retrofit (Android) or Alamofire (iOS).
• Persistence: The single source of truth for data on the device. Data is persisted locally first and then propagated to other components. This enables offline access and reduces network dependency.
• Repository: Mediator between API Service and Persistence. Decouples data sources from the UI.
• Tweet Feed Flow: Components responsible for displaying the scrollable tweet list.
• Tweet Details Flow: Components responsible for displaying a single tweet's details.
• DI Graph: Dependency injection graph. Enables loose coupling and testability (Dagger/Hilt on Android, Swinject/Typhoon on iOS).
• Image Loader: Loads and caches images (Glide/Coil on Android, SDWebImage/Kingfisher on iOS). Check out Image Loading Strategies Deep Dive.
• Coordinator: Manages navigation and flow logic between Tweet Feed and Tweet Details, decoupling the components. Check out Mobile Navigation Architecture Deep Dive.
• App Module: Executable part of the system that "glues" components together. Responsible for application lifecycle management.
• Analytics Service: A module responsible for collecting and transmitting analytics data, such as user actions and performance metrics.

Providing the "Signal"

The interviewer looks for:

• Ability to present the "big picture" without unnecessary details.
• Identification of major building blocks and their interactions.
• Consideration of app modularity and team collaboration.

Frequently Asked Questions

Why is a high-level diagram necessary? Can I skip it?

Not mandatory, but it has advantages:

• Time management: Quickly outlines the system and raises discussion points.
• Modularity: Each component can be a separate module for team collaboration.
• Best practices: Mirrors backend system design and the C4 model for visualizing architecture (https://c4model.com/).

Still not sure? Ask the interviewer if they want a diagram or prefer to jump into specific components.

Deep Dive: Tweet Feed Flow

The interviewer might focus on a specific component, like "Tweet Feed Flow". Discuss:

• Architecture patterns: MVP, MVVM, MVI, Clean Architecture, Redux. MVC is generally discouraged. Choose a well-known pattern for easier onboarding.
• Pagination: Essential for infinite scrolling. (See Pagination section).
• Dependency injection: Improves module isolation and testability.
• Image Loading: Discuss low-res placeholders, full-res loading, and scrolling performance.

Components

• Feed API Service: Abstracts the Twitter Feed API client, fetching paginated data. Injected via DI.
• Feed Persistence: Abstracts cached paginated data storage. Injected via DI.
• Remote Mediator: Fetches next/previous page data and saves it to the persistence layer.
• Feed Repository: Consolidates remote and cached responses into a Pager object using the Remote Mediator.
• Pager: Triggers data fetching from the Remote Mediator and exposes an observable stream of paged data to the UI.
• "Tweet Like" and "Tweet Details" use cases: Implement "Like" and "Show Details" operations. Injected via DI.
• Image Loader: Abstracts image loading from the image loading library. Injected via DI.

Providing the "Signal"

The interviewer assesses:

• Familiarity with common MVx patterns.
• Separation of business logic and UI.
• Understanding of dependency injection.
• Ability to design self-contained modules.

Frequently Asked Questions

How much detail should I provide?

No strict rule. Work with the interviewer; ask if they want more detail or to move on. Observe their body language. If they interrupt, stop and ask if they have questions. Collaboration is key.

Why no mention of specific classes (RecyclerView/UICollectionView) or vendors (Room, CoreData, Realm)?

• To keep the guide platform-agnostic. Libraries evolve rapidly. Abstracting functionality is more robust.
• Vendor selection is subjective and based on experience/trends.
• Large companies often build custom stacks.
• Implementation details are widely available online.

What drawing tool should I use?

Excalidraw and Google Drawings are popular choices. Some prefer collaborative editors and verbal discussion. Privacy policies may limit tool choices.

Client Library Design

If the interview task is to design a Library (e.g., Image Loader, Analytics SDK) rather than an App, the focus shifts to the Public API Design.

Providing the "Signal"

The interviewer assesses:

• API Ergonomics: Do you design interfaces that are intuitive and hard to misuse? (e.g., using Result types instead of Exceptions).
• Stability: Do you consider thread safety, idempotency, and graceful error handling?
• Resource Management: Do you ensure the library doesn't block the main thread or leak memory (avoiding long-lived references to Contexts/Views)?

Check out In-App API Design & Library Architecture Deep Dive for specific patterns (Builder, Singleton, Dependency Injection) and code examples for iOS and Android.

API Design

Cover as much ground as possible. Ask if the interviewer has preferences, or choose an area you know well.

Real-time Notifications

Discuss approaches for real-time updates:

Choose an appropriate approach for the task. For "Design Twitter Feed", a combination of SSE (primary channel for "like" updates and potentially new tweet notifications if scale allows) and Push Notifications (for clients without active connections or less critical updates) could be suitable. Consider WebSockets for features like real-time chat or live video streaming (if in scope).

Protocols

REST (Representational State Transfer)

Text-based, stateless protocol for CRUD (Create, Read, Update, Delete) operations. Relies heavily on HTTP methods (GET, POST, PUT, DELETE).

Check out RESTful API Design Deep Dive for a deeper dive into best practices.

GraphQL

A query language for your API and a server-side runtime for executing those queries. Allows clients to request specific data from multiple resources using a single endpoint.

WebSocket

Full-duplex communication over a single TCP connection.

Learn more:

• WebSocket Tutorial - How WebSockets Work

gRPC (gRPC Remote Procedure Calls)

A modern open source high performance Remote Procedure Call (RPC) framework that runs on top of HTTP/2. Supports bi-directional streaming using a single physical connection.

MQTT (Message Queuing Telemetry Transport)

A lightweight, publish-subscribe network protocol that transports messages between devices. Often used for IoT applications but can be suitable for mobile in specific scenarios.

Select an appropriate protocol. REST is suitable for "Design Twitter Feed" due to its simplicity for many of the initial tasks, though consider GraphQL for more complex data fetching requirements in the future. For truly real-time scenarios (if in scope), explore WebSockets or SSE. gRPC may be overkill for the initial feature set but beneficial if significant performance improvements are needed with more complex features. Consider MQTT only if you specifically need to interact with IoT devices or have very constrained network conditions.

Pagination

Essential for endpoints returning lists. Prevents excessive network and memory usage.

Types of Pagination

Choose an approach. Cursor Pagination is suitable for "Design Twitter Feed" as it decouples pagination from the underlying data store and can handle insertions/deletions reasonably well. However, consider the complexity of implementation. If the dataset doesn't change frequently and the API is read-heavy, Keyset pagination might be a simpler and more performant alternative. Offset pagination is generally discouraged unless the dataset is small and doesn't change frequently.

Example API request:

GET /v1/feed?after_id=p1234xzy&limit=20
Authorization: Bearer <token>
{
  "data": {
    "items": [
      {
        "id": "t123",
        "author_id": "a123",
        "title": "Title",
        "description": "Description",
        "likes": 12345,
        "comments": 10,
        "attachments": {
          "media": [
            {
              "image_url": "https://static.cdn.com/image1234.webp",
              "thumb_url": "https://static.cdn.com/thumb1234.webp"
            },
            ...
          ]
        },
        "created_at": "2021-05-25T17:59:59.000Z"
      },
      ...
    ]
  },
  "cursor": {
    "count": 20,
    "next_id": "p1235xzy",
    "prev_id": null
  }
}

Additional Information

• Evolving API Pagination at Slack
• Everything You Need to Know About API Pagination

Check out Mobile Pagination Strategies & Deep Dive for more details.

Mention HTTP authentication (Authorization header, 401 Unauthorized) and Rate-Limiting (429 Too Many Requests). Keep it brief and simple. Also, discuss potential API versioning strategies if the pagination scheme needs to be updated in the future.

Providing the "Signal"

The interviewer assesses:

• Awareness of network challenges and expensive traffic.
• Familiarity with communication protocols.
• Understanding of RESTful API design.
• Familiarity with authentication and security.
• Understanding of network error handling and rate-limiting.
• Understanding of the trade-offs associated with different pagination strategies.
• Ability to choose a pagination strategy that is appropriate for the specific use case.

Data Storage

Data Storage Options

Local device storage options:

Storage Location

• Internal: Sandboxed, private to the app, and not readable by other apps (with few exceptions). Data is deleted when the app is uninstalled.
• External: Publicly visible and accessible to other apps (subject to permissions). Data persists even after the app is uninstalled. Generally discouraged for sensitive data.
• Media/Scoped: A restricted subset of external storage specifically for media files. Provides enhanced privacy and security compared to traditional external storage. Requires proper permissions and follows specific access patterns.

Storage Type

• Documents (Automatically Backed Up): User-generated data that cannot be easily re-generated (e.g., user profiles, documents, settings). Automatically backed up to cloud storage (e.g., iCloud, Google Drive) unless explicitly excluded.
• Cache: Regeneratable data that can be downloaded again or recreated (e.g., downloaded images, cached API responses). Can be deleted by the system to free up space. Should not contain critical data.
• Temp: Temporary data that is only used for a short period and should be deleted when no longer needed (e.g., temporary files created during processing). Not backed up and can be deleted by the system at any time.

Best Practices

• Store as little sensitive data as possible.
• Use encrypted storage for sensitive data (e.g., EncryptedSharedPreferences, Keychain, or SQLCipher).
• Prevent uncontrolled storage growth. Implement cache cleanup policies and manage file sizes.
• Use appropriate storage locations based on data privacy requirements (internal vs. external storage).
• Consider data backup and restore strategies for user-generated data.
• Use modern data storage APIs (e.g., DataStore on Android) for improved performance and security.
• Avoid storing secrets or API keys directly in the code or configuration files.
• Implement proper error handling and logging for data storage operations.
• Test data storage logic thoroughly to ensure data integrity and prevent data loss.

Approach

Select an approach and discuss pros/cons. For "Design Twitter Feed":

Feed Pagination Table

Use Room (Android) or CoreData (iOS) for storing paginated feed data. The schema could look like this:

// Kotlin (Android - Room)
@Entity(tableName = "feed")
data class FeedItem(
    @PrimaryKey val itemId: String,
    val authorId: String,
    val title: String,
    val description: String,
    val likes: Int,
    val comments: Int,
    val attachments: String, // comma separated list of URLs or JSON
    val createdAt: Date,
    val cursorNextId: String?,
    val cursorPrevId: String?
)

// Swift (iOS - Core Data)
// Add attributes for: itemID, authorID, title, description, likes, comments, 
// attachments, createdAt, cursorNextId, cursorPrevId

• Limit the total number of entries (e.g., 500) to control local storage size. Implement a Least Recently Used (LRU) eviction policy to remove older items.
• Flatten attachments into a comma-separated list of URLs or store them as JSON within a text field. Alternatively, create a separate attachments table and join it with the feed table on itemId for better normalization.
• Store cursorNextId and cursorPrevId with each item to simplify paged navigation. Use nullable types for the cursor IDs.

Attachments Storage

• Store attachments (images, videos) as files in Internal Cached storage.
• Use attachment URLs as cache keys. Implement an ImageLoader library like Coil (Android) or Kingfisher (iOS) to manage downloading, caching, and displaying images.
• Delete attachments when the corresponding feed items are deleted from the feed table. Implement a mechanism for deleting orphaned attachments.
• Limit the cache size (e.g., 200-400 MB) to prevent excessive storage usage. Use an LRU cache eviction policy to remove less frequently used attachments.

For sensitive media data (e.g., private accounts), selectively encrypt image files using encryption keys stored in the Keystore/KeyChain. Use Android Biometrics API or Touch ID/Face ID on iOS for user authentication to access the encrypted data.

Providing the "Signal"

The interviewer assesses:

• Awareness of mobile storage types, security, limitations, and compatibility.
• Ability to design a storage solution for common scenarios, considering performance, security, and scalability.
• Understanding of data modeling and database design principles.
• Familiarity with data persistence libraries and frameworks on Android and iOS.
• Awareness of data encryption and secure storage practices.
• Understanding of cache management and eviction policies.

Additional Topics

Major Concerns For Mobile Development

Key concerns to keep in mind:

• User Data Privacy: Leaks can damage business reputation and result in legal penalties. Ensure compliance with privacy regulations such as GDPR and CCPA.
• Security: Protect against reverse engineering (more important for Android), data breaches, and unauthorized access. Implement security best practices such as data encryption, secure authentication, and authorization.
• Target Platform Changes: New OS releases may limit functionality and tighten privacy. Stay up-to-date with platform changes and adapt the app accordingly. Use compatibility layers or conditional compilation to handle differences between OS versions.
• Non-reversibility of released products: Assume releases are final. Use staged rollouts and server-side feature flags to safely deploy new features and minimize the impact of bugs.
• Performance/Stability:
  • Metered data usage: Cellular data can be expensive. Optimize network requests and minimize data transfer. Use data compression techniques.
  • Bandwidth usage: Constant radio usage drains battery. Batch network requests and use efficient network protocols.
  • CPU usage: High computation drains battery and causes overheating. Optimize algorithms and use background threads for computationally intensive tasks.
  • Memory Usage: High usage increases the risk of background termination. Use memory profiling tools to identify and fix memory leaks.
  • Startup Time: Slow startup creates a poor user experience. Optimize app initialization and lazy-load resources.
  • Crashers/ANRs: UI freezes and crashes lead to poor app ratings. Implement crash reporting and use error tracking tools to identify and fix bugs.
• Battery Consumption: Minimize battery drain by optimizing network requests, CPU usage, and background processing. Use battery profiling tools to identify and fix battery-related issues.
• Geo-Location Usage:
  • Don't compromise privacy with location services.
  • Prefer the lowest possible accuracy. Request increased accuracy progressively.
  • Provide a rationale before requesting location permissions. Use coarse location when fine location is not required. Obfuscate precise locations if possible.
• 3rd-Party SDKs Usage:
  • SDKs can cause performance regressions, security vulnerabilities, and outages.
  • Guard each SDK with a feature flag.
  • Use A/B testing or staged rollout for new SDK integrations.
  • Have a long-term support and upgrade plan. Regularly review and update SDKs to address security vulnerabilities and performance issues.
• App Responsiveness: Ensure the app remains responsive to user input even when performing complex tasks. Use background threads and asynchronous operations to avoid blocking the main thread.
• Testing on Real Devices: Test the app on a variety of real devices to ensure compatibility and performance. Use cloud-based device testing services to access a wide range of devices.
• Licensing and Legal Compliance: Comply with all applicable licensing terms and legal regulations, including open source licenses, privacy policies, and data security laws.

Privacy & Security

Privacy and security should be a central consideration throughout the entire development lifecycle, not an afterthought.

• Minimize user data collection: Only collect data that is absolutely necessary for the functionality of the app.
  • Avoid device identifiers (use resettable, pseudo-anonymous IDs). Consider using Instance IDs instead.
  • Anonymize collected data whenever possible. Aggregate data and remove personally identifiable information (PII).
  • Explain what data you are collecting, how it is used, and who it is shared with, in easy to understand terms.
  • Provide users with transparency and control over their data. Allow users to access, modify, and delete their data.
  • Implement data retention policies to ensure that data is only stored for as long as it is needed.
• Minimize permission usage: Request only the permissions that are necessary for the app's functionality.
  • Be prepared for permission denials and respect the user's choice. Handle permission denials gracefully and provide alternative functionality if possible.
  • Be prepared for auto-reset permissions. Implement a mechanism for requesting permissions again if they are auto-reset by the system.
  • Be prepared for app hibernation of unused apps.
  • Delegate functionality to system apps (Camera, Photo Picker, File Manager) to reduce the need for permissions.
• Secure Data Storage:
  • Assume on-device storage is not secure, even when using KeyStore/KeyChain functionality. Implement encryption to protect sensitive data.
    • Encrypt all sensitive data at rest using strong encryption algorithms.
    • Store encryption keys securely in the KeyStore/KeyChain or a hardware security module (HSM).
    • Use secure data storage APIs such as EncryptedSharedPreferences on Android and the Keychain on iOS.
    • Implement proper access controls to restrict access to sensitive data.
  • Assume backend storage is also not secure.
    • Use end-to-end encryption to protect data in transit and at rest. Implement server-side encryption to protect data stored in the backend.
    • Use secure communication protocols such as HTTPS.
    • Implement proper authentication and authorization mechanisms to restrict access to backend resources.
• Secure Communication:
  • Use HTTPS for all network communication.
  • Validate server certificates to prevent man-in-the-middle attacks.
  • Implement proper authentication and authorization mechanisms.
  • Use secure data transfer protocols.
• Code Security:
  • Protect against reverse engineering by obfuscating the code and using anti-tampering techniques.
  • Validate user input to prevent injection attacks.
  • Use secure coding practices to avoid common security vulnerabilities.
  • Perform regular security audits and penetration testing.
• Privacy by Design:
  • Implement privacy considerations at every stage of the development process.
  • Conduct a Data Protection Impact Assessment (DPIA) to identify and mitigate privacy risks.
  • Use privacy-enhancing technologies (PETs) such as differential privacy and homomorphic encryption.
• Regular Security Updates:
  • Stay up-to-date with the latest security vulnerabilities and patches.
  • Regularly update dependencies and third-party libraries.
  • Implement a security incident response plan.
• Privacy Compliance:
  • Comply with all applicable privacy regulations, such as GDPR and CCPA.
  • Obtain user consent before collecting or processing personal data.
  • Provide users with the right to access, modify, and delete their data.
  • Implement data breach notification procedures.
• Transparency and User Education:
  • Assume platform security/privacy rules will change. Control critical functionality with remote feature flags to quickly adapt to new regulations.
  • User perception of security is crucial. Discuss how you would educate users about data collection, storage, and transmission.
    • Be transparent about data collection practices.
    • Provide users with clear and concise information about how their data is used.
    • Educate users about security threats and best practices.
    • Implement a user-friendly privacy policy.

Cloud vs On-Device

Choose between running functionality on a device or in the cloud. Especially relevant for On-Device AI, Machine Learning, and complex processing.

Cloud Execution

On-Device Execution

Things you should never run on a device:

• New "resource" creation (sensitive): Generating coupons, tickets, promo codes, or other sensitive resources should always be done server-side to prevent fraud.
• Transactions and Payment verification: Never process payment information directly on the device. Delegate to a 3rd-party SDK (e.g., Stripe, Braintree) or redirect the user to a secure payment gateway. Sensitive financial calculations should be performed on the server.
• Complex Business Logic or sensitive Algorithms: Sensitive algorithms, user authentication, or core business rules should be kept on the server for security and control. This protects valuable IP and allows for updates without app releases.

Offline and Opportunistic State

Offline state ensures app usability without a network connection:

• User can like/comment/delete tweets offline.
• UI updates "opportunistically" assuming requests will be sent when online.
• App displays cached results.
• App notifies users about its Offline State, and any pending actions.
• State changes are batched and sent when the network goes back online.

Check out Offline-First Architecture & Data Synchronization Deep Dive for a deep dive into synchronization strategies and conflict resolution.

Request de-duplication

Ensure retrying requests won't create duplicates on the server (idempotence). Use unique client-side generated request IDs and server-side de-duplication. The server should store a record of processed request IDs for a certain period.

Use the Idempotency-Key HTTP header to ensure a client can safely retry a request without accidentally performing the same operation twice. The client generates a unique key for each request and includes it in the header. If the server receives the same key again, it knows that the request has already been processed and can return the previous response.

Example (Kotlin):

// Client-side request
val idempotencyKey = UUID.randomUUID().toString()
val request = Request.Builder()
    .url("https://example.com/api/like")
    .header("Idempotency-Key", idempotencyKey)
    .post(requestBody)
    .build()

Synching Local and Remote States

Handle state synchronization across multiple devices. Implementing conflict resolution is crucial.

Conflict Resolution Strategies (on the server, using Remote Conflict Resolution):

• Last Write Wins with Timestamps (Not recommended in most cases): The server simply accepts the latest update based on the timestamp. This is the simplest approach, but it can lead to data loss if clocks are not synchronized or if updates are received out of order.
• Conflict Detection and Resolution UI: The server detects a conflict and returns an error to the client, prompting the user to manually resolve the conflict through a UI. This approach provides the most control over conflict resolution but can be cumbersome for the user.
• Operational Transformation (OT): A more sophisticated approach that transforms operations based on the history of changes. This allows for concurrent edits to be merged seamlessly. OT is more complex to implement but provides a better user experience.
• Conflict-Free Replicated Data Types (CRDTs): Data structures that are designed to be merged automatically without conflicts. CRDTs are a good choice for applications where data consistency is critical.

More Information

Offline functionality for Trello’s mobile applications: Airplane Mode: Enabling Trello Mobile Offline

Caching

Caching is essential for offline support, reducing network usage, and improving performance. It involves storing data in memory (for immediate access) or on disk (for persistence).

Check out Caching Strategies Deep Dive for details on strategies (Read-Through, Stale-While-Revalidate), eviction policies, and security.

Quality Of Service (QoS)

Implementing Quality of Service (QoS) for network operations is critical for providing a smooth user experience while conserving device resources. It involves prioritizing network requests based on their importance (User-Critical vs Background).

Check out Quality of Service Deep Dive for details on concurrency limits, request prioritization, and power management.

Resumable Uploads (Chunked Uploads)

Resumable uploads break down large files into smaller chunks, allowing the upload to be paused and resumed. This improves reliability on unstable networks.

Check out Resumable Uploads Deep Dive for the initialization process, chunk appending, and advantages/disadvantages.

Prefetching

Prefetching proactively fetches data before it is explicitly requested (e.g., preloading the next page of a feed). This reduces latency and improves UX but consumes more resources.

Check out Prefetching Deep Dive for types of prefetching, implementation strategies, and how to mitigate negative impacts.

Conclusion

Navigating system design interviews can feel like an overwhelming task. Remember that there's inherent randomness – the process can differ significantly across companies and even between interviewers within the same company. The key is to focus on what you can control and learn from what you can't.

Things You Can Control:

• Your Mindset: Approach each interview with a positive and curious attitude. Be enthusiastic about solving the problem and demonstrate a willingness to learn. Be respectful and collaborative, even under pressure. A good attitude leaves a lasting impression.
• Your Preparation: Thorough preparation is your strongest weapon. Practice mock interviews with peers, mentors, or online services. The more you practice, the more comfortable and confident you'll become.
• Your Knowledge Base: Continuously expand your knowledge of mobile technologies, architectural patterns, and design principles. Stay updated with the latest trends and best practices in iOS and Android development.
  • Immerse yourself in open-source projects: iOS, Android
  • Curated List of Real-world Design Blog Posts
• How you articulate your thought process: Clearly and concisely explain your reasoning behind design choices, and be prepared to justify your decisions with evidence and trade-offs. Walk the interviewer through your process step-by-step.
• Your Resume: Craft a compelling resume that highlights your accomplishments and demonstrates the impact you've made in your previous roles. Quantify your achievements whenever possible. Tailor your resume to match the specific requirements of the position you're applying for.

Things Beyond Your Control:

• The Interviewer's Style and Bias: You might encounter interviewers with different personalities, preferences, or even unconscious biases. Focus on presenting your best self and don't take any perceived negativity personally.
• The Competition: There might be other highly qualified candidates with more experience or skills. Focus on showcasing your unique strengths and value proposition.
• The Hiring Committee's Decision: The final decision rests with the hiring committee, which considers various factors beyond your control, such as budget constraints, team dynamics, and organizational priorities.

Embrace the Process, Learn from Every Experience:

Remember that every interview, regardless of the outcome, is a valuable learning opportunity. Analyze your performance, identify areas for improvement, and seek feedback from others. Don't be discouraged by rejections – view them as stepping stones toward your ultimate goal.

Failure is a part of the process. Even the most experienced engineers face rejection at some point in their careers. What matters is how you respond to setbacks. Use each experience to refine your skills, expand your knowledge, and strengthen your resolve.

Keep practicing, keep learning, and keep pushing yourself. The right opportunity will eventually come your way. Believe in your abilities, and never give up on your dreams. Good luck on your journey!

Frequently Asked Questions

Q: How do you know this approach works? Why is this guide necessary?

• A: There's no one-size-fits-all approach to system design interviews. The structure of the interview heavily depends on the interviewer's style and the company's specific needs. However, having a structured framework like this helps both the candidate and the interviewer focus on the content of the discussion rather than getting bogged down in the organizational aspects of the interview. It provides a common language and a starting point for exploration.

Q: Can you go a bit deeper into [specific technology/pattern/component]?

• A: This guide aims to provide a broad overview of key concepts. It doesn't delve into extreme detail on any specific technology or implementation. The goal is to equip you with a foundational understanding, encouraging you to draw upon your personal experience and expertise to provide concrete solutions during the interview. Remember, the interviewer is assessing your problem-solving skills and your ability to adapt and apply your knowledge.

Q: I'm an interviewer - won't this guide encourage candidates to memorize solutions and "cheat" during the interview?

• A: System design is far more nuanced than coding challenges. Memorizing a specific solution is rarely sufficient, as interviewers can easily adapt the problem or introduce new constraints. This guide aims to provide a foundational understanding, but success hinges on a candidate's ability to reason critically, apply their knowledge creatively, and articulate their thought process clearly. It's usually quite apparent when a candidate is relying solely on memorization rather than demonstrating genuine understanding. Exposure to common patterns and approaches can reduce interview anxiety and allow the candidate to showcase their problem-solving skills more effectively.

Q: What if the interviewer wants to focus on a technology I'm not very familiar with?

• A: Be honest about your strengths and weaknesses. If the interviewer veers into an area where your knowledge is limited, acknowledge that and explain your general understanding of the concepts involved. Emphasize your willingness to learn and your ability to quickly acquire new knowledge. Suggest alternative solutions using technologies you are more comfortable with.

Q: How important is it to draw a diagram? What if I'm not a visual thinker?

• A: Visualizing the system architecture through diagrams can be extremely helpful for both you and the interviewer. It provides a clear and concise way to communicate the overall design and the interactions between components. If you're not naturally inclined to draw diagrams, practice beforehand! There are many online tools that can help. However, if you feel a diagram is not the best way to express yourself, communicate clearly why and offer an alternative representation, such as a well-structured verbal explanation.

Q: How should I handle disagreements with the interviewer?

• A: It's perfectly acceptable to have different opinions than the interviewer. However, approach the discussion respectfully and professionally. Clearly articulate your reasoning and be willing to consider alternative viewpoints. Avoid being argumentative or dismissive. If you strongly disagree, you can politely acknowledge the interviewer's perspective while reiterating your own rationale. The goal is to demonstrate your critical thinking skills and your ability to engage in constructive dialogue.

Q: What if I get stuck and can't think of a solution?

• A: Don't panic! It's okay to pause and take a moment to gather your thoughts. Communicate your thought process out loud. Explain what you're considering and what challenges you're facing. Ask clarifying questions to help you better understand the problem. If you're still stuck, ask the interviewer for a hint or guidance. The interviewer is more interested in seeing how you approach a difficult problem than in finding the perfect solution.

Q: How much detail should I provide when discussing specific components?

• A: The level of detail should depend on the interviewer's interest and the time available. Start with a high-level overview and then dive deeper into specific aspects as prompted by the interviewer. Pay attention to their body language and ask if they want you to elaborate or move on to another topic. Don't get bogged down in minute details unless the interviewer specifically requests it.

Q: How do I choose between different architectural patterns (MVP, MVVM, MVI, etc.)?

• A: There's no single "best" architectural pattern. Each pattern has its own advantages and disadvantages, and the best choice depends on the specific requirements of the project. Be prepared to explain the pros and cons of different patterns and justify your choice based on factors such as complexity, testability, and maintainability. Familiarity with common patterns such as MVVM with a unidirectional data flow (like MVI or Redux) is generally favored.

Q: How can I improve my communication skills for system design interviews?

• A: Practice explaining complex technical concepts in a clear and concise manner. Use diagrams and visual aids to illustrate your ideas. Practice active listening and ask clarifying questions to ensure that you understand the interviewer's perspective. Seek feedback from others on your communication style. Record yourself explaining technical concepts and analyze your performance. Focus on being clear, concise, and confident in your communication.

Q: Should I memorize specific code snippets or API calls for the interview?

• A: While having a general understanding of common APIs and libraries is helpful, it's generally not necessary (or advisable) to memorize specific code snippets. The focus should be on demonstrating your understanding of the underlying concepts and your ability to apply them to solve problems. If you need to refer to a specific API or code example, explain why you're using it and what it accomplishes.

Q: Is it better to focus on a single area of mobile development (e.g., iOS or Android) or have broad knowledge across both platforms?

• A: The ideal approach depends on the specific role and the company's needs. Some roles may require deep expertise in a single platform, while others may value broad knowledge across both platforms. Be honest about your skills and experience. If you have broad knowledge, highlight your ability to learn new technologies and adapt to different platforms. If you have deep expertise in a single platform, emphasize your ability to contribute immediately to the team and your willingness to learn about other platforms as needed.

Additional Information and Resources

Explore these resources to further enhance your preparation:

More System Design exercises can be found here!

Level-Specific Expectations in System Design Interviews

The focus and complexity of system design interviews will vary depending on the target engineering level. The key is to demonstrate skills and thinking appropriate for that level. An approximate engineering level breakdown can be found here (though leveling structures can differ between companies).

NOTE: Years of experience are only a rough guide and don't directly translate to a specific level. Individual skills, contributions, and impact weigh much more heavily.

Junior Engineer

• The system design round is often optional.
• Focus is on foundational knowledge, understanding basic concepts, and implementing small, well-defined components.
• Expect questions about data structures, algorithms, and basic design patterns.
• Limited or no experience with large-scale system design is expected.
• Demonstrate a willingness to learn and a capacity for growth.

Mid-Level Engineer

• The system design round emphasizes practical implementation and the application of architectural patterns.
• Expect discussions about building specific components using platform-specific libraries and frameworks.
• Demonstrate proficiency in coding, testing, and debugging.
• Ability to translate high-level requirements into concrete implementation details.
• Understand the trade-offs between different implementation approaches.

Senior Engineer

• The system design round shifts to a more high-level perspective.
• Focus on system architecture, component interactions, and scalability.
• Expect discussions about multiple components, their communication, and overall system behavior.
• Implementation details are less critical, unless they significantly impact performance.
• Ability to select appropriate technologies and justify the choices.
• Understand the trade-offs between different architectural patterns and design choices.
• Think about modularity, testability, and maintainability.

Staff Engineer

• The system design round transitions from technical to strategic considerations.
• Expect discussions about the target audience, available resources (computational and human), expected traffic, and deadlines.
• Prioritize business needs over technical implementation details.
• Ability to explain how to reduce time-to-market, safely roll out features, and handle large-scale outages.
• Deep understanding of user privacy and its legal implications.
• Consider the long-term impact of design decisions.
• Focus on building consensus and influencing technical direction.
• Expert in scaling systems, choosing the right databases, and understanding networking principles.
• Ability to identify and mitigate risks.
• Think about cost optimization and resource allocation.

Looking for more content?

Interview Template

Use the TEMPLATE.md to structure your notes during mock interviews.

System Design Exercises

Check out the collection of mobile system design exercises.

Common System Design Interview Mistakes

Check the guide here.

Mock Interviews

Check out the mock interview archive or sign-up to be a candidate yourself!

Consider "starring" the repository to help other people discover the guide! Thank you!