← Back to Directory

Anthropic interview notes, page 27 of 32

Mid-sized

Most of these Anthropic write-ups come from Software Engineers (277), then Machine Learning Engineers (31) and Product Managers (10). The most reported rounds are the technical phone screen (129), the system design round (86), the behavioral / leadership round (45) and the hiring manager screen (17).

319 interview notes · updated Sep 2026

Anthropic·Software Engineer·Technical Phone Screen

Jun 2026
Interviewed for a software engineering role at Anthropic and got a coding problem centered on batch image processing. The question had a practical systems flavor to it, less about clever tricks and more about whether you think about I/O, parallelism, and clean interfaces from the start.
  • Given m input images and n processing pipelines (each an ordered sequence of operations like resize, rotate, crop, etc.), write a function that applies every pipeline to every image and outputs all m*n results efficiently. The interface should read images from file paths and write outputs to a directory with deterministic naming.
  • How would you scale this image processing system for very large values of m and n, and what parallelization strategy would you choose?

“My first instinct was to just nest two loops and call it a day, and they let me write that out before asking how I'd handle scale.”

View Post

Anthropic·Software Engineer·Onsite - Behavioral / Leadership

Jun 2026
Behavioral round at Anthropic for a software engineering role. Just one question but it required a real story, not a generic conflict answer.
  • Tell me about a time you disagreed with someone and had to change your approach to bring them around.

“I had a decent story ready but the part I fumbled was explaining why I changed my approach rather than just what I changed.”

View Post

Anthropic·Software Engineer·Technical Phone Screen

Jun 2026
Technical phone screen for a software engineering role at Anthropic. One meaty parsing question that went pretty deep into tree construction and edge case handling. Not a vibe check, they wanted actual working code.
  • Given a multi-line stack trace string from a single thread, design and implement a parser that extracts frames in order (most recent to oldest), reconstructs the call hierarchy as a tree, identifies the deepest frame tied to the thrown exception, and correctly handles nested 'Caused by:' sections, abbreviated '... N more' lines, and malformed or missing lines. Walk through your data structures, time/space complexity, and provide working code.

“This one took me a second to even parse the problem statement, which is ironic.”

View Post

Anthropic·Software Engineer·Technical Phone Screen

Jun 2026
Interviewed for a software engineering role at Anthropic and got a classic systems design-meets-coding hybrid question. Nothing too exotic, but the follow-up discussion on trade-offs is where it got interesting.
  • Implement the core operations of a task management system: adding a task, fetching a task by ID, and updating its fields. Then walk through your choice of in-memory data structure and the trade-offs involved.

“Started with a HashMap keyed on task ID, which felt obvious, and they let me run with it.”

View Post

Anthropic·Software Engineer·Technical Phone Screen

Jun 2026
Anthropic SWE interview where they handed me a broken GRPO training implementation and asked me to find three bugs. Pretty heads-down technical, no fluff.
  • You're given a partial GRPO training implementation with exactly three bugs. Identify and fix each one.

“The bugs were all in the core training logic.”

View Post

Anthropic·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Interviewed for an ML engineer role at Anthropic. Just one question to go off of here, but it was the kind of thing that sounds casual until you're actually sitting there trying to not sound like you just read a few tweet threads.
  • How do you keep up with new developments in machine learning?

“I rambled a bit.”

View Post

Anthropic·Software Engineer·Recruiter / HR Screen

Jun 2026
Interviewed for a hardware engineer role at Anthropic. One question, pretty short interaction, felt more like a values screen than anything technical.
  • How much do you care about safety and security?

“Pretty loaded question for a hardware role, and I kind of fumbled it at first by going straight into physical security and tamper resistance.”

View Post

Anthropic·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Anthropic ML interview, one meaty technical question on probability calibration that covered a surprising amount of ground for a single prompt. Felt like a design question dressed up as a theory question.
  • You have a binary classifier whose raw output scores aren't well-calibrated. Walk through how you'd fix that using Platt scaling and isotonic regression, explain the tradeoffs between the two, and describe how you'd actually measure whether your calibration is any good.

“I knew Platt scaling well enough, it's just fitting a logistic regression on top of your model's scores using a held-out set.”

View Post

Anthropic·Software Engineer·Technical Phone Screen

Jun 2026
Interviewed at Anthropic for a SWE role and got a classic systems problem that sounds deceptively simple until you're actually in it.
  • Given a file system, how would you find all duplicate files?

“My first instinct was to just hash everything and compare, which is fine but I jumped to it too fast without talking through the tradeoffs.”

View Post

Anthropic·Software Engineer·Onsite - System Design / Architecture

Jun 2026
System design round at Anthropic for a software engineer role. One meaty question that sprawled into a full distributed systems conversation, covering everything from topology choices to handling stragglers. Left feeling like I could've gone deeper on the math.
  • You need to distribute a large file from cloud storage to 1000 servers as fast as possible. The uplink from cloud storage is capped at 1 Gb/s and each server's NIC is also capped at 1 Gb/s. Walk through your design, covering tree vs chain vs peer-to-peer distribution, the theoretical minimum completion time, chunking, and pipelining.
  • How does your distribution design change when some of the 1000 hosts have slow or unreliable network connections? How do you prevent slow nodes from holding up the rest of the fleet?

“This one sprawled in a good way.”

View Post