Airbnb DevOps / SRE Interview Questions

1.Describe a time when you had to design or implement infrastructure that needed to work reliably across different regions with varying network conditions and local regulations.

2.Tell me about a time when you had to ensure that infrastructure or monitoring tools were accessible and usable by team members with different technical backgrounds or experience levels.

3.Tell me about a time when improving monitoring or alerting required significant infrastructure changes that your team initially pushed back on. How did you convince them it was worth the effort?

4.Describe a time when you pushed for better deployment practices or infrastructure tooling that would improve developer experience, even though it required significant upfront investment.

5.Walk me through a critical incident or outage where your team members were distributed across multiple time zones. How did you coordinate the response and what did you learn?

6.Tell me about a time when you had to make infrastructure decisions that affected both host tools and guest-facing features differently. How did you balance these competing needs?

7.You're deploying a new feature to our booking service that processes millions of reservations daily. The deployment fails with a 500 error rate spike to 15% and you have 30 seconds to decide whether to rollback or debug forward. Walk me through your decision process.

8.We're migrating our listing management service from a Rails monolith to microservices on Kubernetes. During migration, we need zero downtime and the ability to route different host operations to either the old or new system. Design the migration strategy and infrastructure.

9.Our payments microservice needs to handle currency conversion for listings in 190+ countries. The current architecture queries an external FX API synchronously on every booking, which is causing timeouts. How would you redesign this?

10.We want to A/B test a new search ranking algorithm for Stays, but the current setup requires manually deploying different Docker containers for each variant. Design a system that allows our data scientists to deploy ranking experiments without engineering involvement.

11.You notice that our Kafka cluster handling booking events is experiencing message lag during peak booking hours. Some downstream services are processing events 10+ minutes late, affecting host notifications and guest confirmations. How would you investigate and resolve this?

12.Our Trust & Safety team needs to run ML models that analyze listing photos, guest reviews, and host verification data in real-time during booking attempts. The current batch processing approach takes 6+ hours. Design an architecture that provides risk scores within 500ms of booking submission.

13.Design a monitoring system that can detect when our search service is returning significantly fewer available properties than expected for popular destinations like Paris or Tokyo. The system should alert within 5 minutes when availability drops below historical baselines.

14.Our Experiences platform needs to handle booking confirmations for activities that have limited capacity and specific time slots. Design a system that prevents overbooking while ensuring hosts can update availability in real-time across multiple time zones.

15.We want to implement circuit breaker patterns across our microservices architecture, but different services have very different traffic patterns and failure modes. Design a configurable circuit breaker service that can adapt to each microservice's characteristics without manual tuning.

16.Design a secrets management system for our Kubernetes clusters that can handle database credentials, API keys, and certificates. The system needs to work for both application secrets and infrastructure automation, with automatic rotation for high-risk credentials.

17.Our listing photos are processed through multiple ML pipelines for quality scoring, safety detection, and search indexing. Currently each pipeline processes photos independently, leading to duplicate work and inconsistent results. Design a unified photo processing architecture that eliminates redundancy while maintaining pipeline isolation.

18.Design an infrastructure provisioning system that can spin up complete environments for feature development, including all microservices, databases, and external service mocks. The system should support our 'Live and Work Anywhere' policy by allowing engineers to create environments from any location with varying network conditions.

19.Tell me about a time when you had to convince multiple engineering teams to adopt a new operational practice or tool that would benefit the overall platform but required each team to change their workflows.

20.Describe a situation where you had to balance automation and manual intervention in your operational processes. What factors influenced your decision on where to draw that line?

21.Walk me through a time when you identified that your monitoring or alerting was creating alert fatigue for your team. How did you approach fixing this while ensuring critical issues weren't missed?

22.Tell me about a time when you had to lead an incident response that involved coordinating with product teams, support teams, and potentially external partners. How did you manage the communication and decision-making process?

23.Describe a situation where you identified that your team's on-call practices or operational processes were preventing team members from doing their best work. What changes did you implement and how did you measure success?

24.Tell me about a time when you had to influence engineering leadership to prioritize reliability or operational improvements over new feature development. How did you build the case and what was the outcome?

25.Estimate how many database queries our search service executes per minute during peak booking hours for popular destinations like New York or London. Walk me through your reasoning.

26.Our host onboarding completion rate dropped 8% last month across all regions. No code changes were deployed. How would you investigate this as an SRE?

27.Design a capacity planning model for our booking confirmation service that handles currency conversion, payment processing, and host notifications across 190+ countries during peak seasons.

28.Estimate the infrastructure cost impact if we reduced our booking API response time from 800ms to 400ms. Include both the engineering effort and ongoing operational costs.

29.You're tasked with building a system to automatically scale our listing photo processing pipeline during major host onboarding campaigns like 'Summer Release' promotions. The system needs to handle 10x normal photo upload volume while maintaining quality standards.

30.Design a real-time alerting system that can detect when booking success rates drop below normal patterns, accounting for the fact that different property types and regions have vastly different baseline conversion rates.