Technology Selection & Deep Technical Knowledge Questions

Deep understanding of specific technologies relevant to complex system design. Master databases (PostgreSQL, Cassandra, DynamoDB, Elasticsearch), message queues (Kafka, RabbitMQ), caching systems (Redis), search engines, and frameworks. Understand their strengths, weaknesses, trade-offs, operational characteristics, scaling patterns, and common pitfalls. Be able to justify technology choices based on specific system requirements.

MediumSystem Design

45 practiced

Design a caching strategy for ML model inference that balances freshness and latency. Requirements: 5ms target latency, features update every 1–5 minutes, occasional feature recomputations. Describe cache placement (client, edge, Redis), key structure, TTL strategy, cache invalidation on update, and how you prevent stale or inconsistent predictions.

MediumTechnical

47 practiced

Write a PostgreSQL query using window functions to compute 7-day user retention: for each cohort (user's first event date), show the number of users who return on each day 0..6. Provide a sample schema and example rows, then write the SQL and briefly explain how it works.

HardTechnical

76 practiced

You're tasked with scaling PostgreSQL to support both transactional workloads and heavy analytical queries. Compare vertical scaling, read replicas, logical replication into an analytics cluster, and sharding via Citus. For each approach discuss operational complexity, query capabilities (joins across shards), maintenance, and suitability for ML workloads that need fresh but heavy aggregations.

EasyTechnical

39 practiced

You must choose a partition key for storing IoT telemetry in a wide-column store (Cassandra/DynamoDB). Queries: (1) get last 24h of telemetry for a single device; (2) compute daily aggregates per device; (3) occasionally query across devices for a time range. Propose a partition-key and clustering key strategy that reduces hot partitions and supports the query types. Explain trade-offs and mitigation techniques for hot keys.

HardTechnical

54 practiced

For telemetry at 500k writes/sec, compare wide-column stores (Cassandra/Scylla), document stores (MongoDB), and key-value stores (DynamoDB) in terms of write scalability, compaction, consistency, ability to run aggregations, and operational overhead. Which would you recommend and why? Include mention of data modeling patterns to keep throughput high.

Unlock Full Question Bank

Get access to hundreds of Technology Selection & Deep Technical Knowledge interview questions and detailed answers.

Join thousands of developers preparing for their dream job.