Anthropic Interview Questions — Page 22

Anthropic·Software Engineer·Onsite - System Design / Architecture

Jun 2026

System design round at Anthropic for a software engineer role. The whole session was basically one massive question about distributed statistics at scale, which sounds straightforward until you're 20 minutes in and realizing you forgot to talk about fault tolerance.

You have hundreds of billions of records spread across many machines. Design a system to compute the global median and global mode. Cover data partitioning, local summaries, how reducers combine partial results, and how you keep network I/O low. Compare exact vs approximate methods, handle skew and faults, support incremental updates, and extend the design to a sliding-window version over an unbounded stream.

“This is a lot of surface area for one question.”

View Post

Anthropic·Product Manager·Onsite - Product Sense / Strategy

Jun 2026

Interviewed for a PM role at Anthropic and got hit with a question about anthropomorphizing LLMs that I did not see coming from that angle.

What are the risks of assuming that large language models think or feel the way humans do?

“I fumbled the opener because I started listing obvious UX stuff like users over-trusting the model, but they kept pushing and I realized pretty late they wanted something more fundamental.”

View Post

Anthropic·Software Engineer·Technical Phone Screen

Jun 2026

Anthropic coding round, got a multithreading problem that looked deceptively like a standard BFS until you realize half the complexity is just keeping things thread-safe and not deadlocking yourself into a corner.

Implement a multithreaded web crawler. Given a start URL and a function that returns all URLs reachable from a page, crawl every reachable URL sharing the same hostname. Each URL should be visited at most once, and you must use multiple threads to fetch pages concurrently.

“I started with the hostname filtering which felt easy, just split on slashes and compare.”

View Post

Anthropic·Software Engineer·Onsite - System Design / Architecture

Jun 2026

Interviewed for a SWE role at Anthropic and got a system design question focused on batching infrastructure for LLM inference. Pretty domain-specific stuff, not your typical distributed systems question.

Design an end-to-end batching system for LLM queries, including an API that efficiently groups and processes sampling requests at scale.

“This one took me a minute to even figure out what angle they wanted.”

View Post

Anthropic·Software Engineer·Technical Phone Screen

Jun 2026

Anthropic coding screen for a software engineer role. One problem, pretty focused on data structures and binary search. The problem itself was clean but the O(log n) constraint is where things get interesting.

Design a time-aware key-value store that supports setting a key with a timestamp and retrieving the value that was active at or before a given timestamp. GET queries must run in O(log n) time relative to the number of writes for that key.

“The core idea clicked fast: store each key's history as a sorted list of (timestamp, value) pairs and binary search on GET.”

View Post

Anthropic·Software Engineer·Technical Phone Screen

Jun 2026

A technical interview at Anthropic for a software engineering role, centered on a pretty gnarly open-ended optimization problem. The whole thing felt less like a coding screen and more like a take-home design exercise compressed into a conversation.

You're handed a compute kernel and a cycle-accurate simulator. You have two weeks to maximize speedup without changing the outputs. Walk through your complete approach from baseline to final result.
How would you validate that your optimizations haven't broken correctness after each change, and how do you make sure you're not just overfitting to the simulator's behavior?
Pick the first three optimizations you'd try and explain why you'd prioritize those over everything else.

“This one is deceptively open.”

View Post

Anthropic·Software Engineer·Technical Phone Screen

Jun 2026

Anthropic software engineer interview with a systems-flavored coding problem. The main question was a file deduplication tool, followed by a pretty deep follow-up discussion on scaling and parallelism. Nothing behavioral, just technical the whole way through.

Implement a file deduplication tool that recursively scans a directory, groups files by size, uses content hashing to find duplicates, and returns only groups with at least two matching files. Must handle large files without loading them fully into memory and deal gracefully with permission errors or files changing mid-scan.
Is this workload I/O-bound or CPU-bound, and how would you figure out which one it is?
How would you handle deduplication across millions or billions of files?
How would you parallelize the directory scan and the hashing stages?
How would you support near-realtime duplicate detection as files are created or modified?

“This one took me a minute to get into the right headspace.”

View Post

Anthropic·Software Engineer·Onsite - Multi Round

Jun 2026

Interviewed at Anthropic for what looked like a senior technical or research role. The whole thing was basically a deep-dive presentation on your own work, which sounds chill until you realize the follow-up questions are where they actually evaluate you.

Prepare and deliver a 20-30 minute presentation on a recent project you led, covering the problem, your technical approach, key decisions and trade-offs, results, and what you'd do differently. Be prepared for deep follow-up questions.

“This format is deceptively hard.”

View Post

Anthropic·Software Engineer·Onsite - System Design / Architecture

Jun 2026

System design round at Anthropic for a software engineer role, focused entirely on one deep problem: making an in-memory LRU cache survive crashes and restarts. The question had a lot of layers and the hints they gave mid-interview made it clear they wanted you to reason through trade-offs, not just recite a solution.

You have an in-memory LRU cache with O(1) get/put. Right now it's fully volatile. Design a persistence strategy so the cache can recover its state after a process crash or restart, including what to persist, how to reconstruct LRU ordering, whether to use snapshots or a write-ahead log or both, and how to handle partial writes during recovery.
The cache logic runs at microsecond speed but disk flushes take milliseconds. How do you prevent persistence from blocking live get/put traffic, and where exactly do you draw the line between synchronous and background work?
How would you take a snapshot of roughly 1 GB of cache state without stalling live traffic for the several seconds it would take to write?
If exact LRU eviction order after recovery were a hard requirement rather than approximate warmth, what would you change and what would it cost you?
How does the persistence and recovery story change if you extend this to a sharded and replicated cache, where each shard has its own WAL and a follower needs to stay in sync?

“This took me a while to even frame correctly.”

View Post

Anthropic·Software Engineer·Onsite - System Design / Architecture

Jun 2026

System design round at Anthropic for a software engineering role. The whole session was basically one deep question about file deduplication that kept branching into new territory every time I thought I'd covered it.

Design a file deduplication system for a large-scale storage platform. Cover chunking strategies, hash function selection, collision handling, memory constraints for large files, streaming ingestion, parallelization, I/O optimization, and the data structures you'd use for fast lookups.

“This question ate the entire interview.”

View Post