← Back to Directory

Sesame

Small

Screna has 6 Sesame interview write-ups. The most reported round is the technical phone screen (3).

6 interview notes · updated Jun 2026

Sesame·Software Engineer·Technical Phone Screen

Jun 2026
Got a technical phone screen for a software engineer role at Sesame AI that was basically one meaty design question the whole time. Interesting problem, more nuanced than it looks at first glance.
  • Design a StreamingMetrics class that ingests ASR (speech recognition) events, each containing a username and word count, and tracks two things: which user produced the longest single sentence, and how many unique users have been seen. The constructor takes a lookback window size limiting how many recent events are kept in memory. The ingest method must run in O(1) time and memory must stay bounded.

“This one took me a minute to fully unpack.”

View Post

Sesame·Software Engineer·Technical Phone Screen

May 2026
Sesame software engineer interview with a low-level systems design problem. The question was deceptively concrete but opened up into a pretty broad design discussion once you got past the initial implementation.
  • You have an in-memory filesystem backed by a fixed-size list of nodes, where each node is a fixed-size list of characters. Implement a print method that correctly traverses and outputs the stored content, and explain how you'd handle file boundaries within or across nodes, partially-filled nodes, and how the design extends to read/write operations.

“This one looked simple at first glance.”

View Post

Sesame·Machine Learning Engineer·Technical Phone Screen

May 2026
Sesame AI had me implement a ring buffer from scratch, which sounds like a classic CS exercise until you're actually in it and realize how many edge cases there are. The technical depth expected here felt more systems-level than I expected for an ML role.
  • Implement a circular buffer with a fixed byte capacity. It should support writing bytes (with some overflow policy), reading up to N bytes from the front, and reporting how many bytes are currently readable. Walk through how you handle wraparound and edge cases, and briefly discuss thread safety.

“The wraparound part I got fine, modulo arithmetic is straightforward once you draw it out.”

View Post

Sesame·Software Engineer·Onsite - System Design / Architecture

May 2026
System design round at Sesame for a software engineering role. The whole thing was basically one big networking question dressed up as a product feature, which I didn't fully appreciate until I was halfway through explaining threads.
  • Design a TCP-based chat server where multiple clients can connect, send messages that get broadcast to all other connected clients, and the server handles clients joining and leaving cleanly. Walk through your concurrency model, how you'd handle message framing on a raw TCP stream, and graceful shutdown.

“I jumped straight into thread-per-connection because it felt obvious, and the interviewer just kind of waited.”

View Post

Sesame·Software Engineer·Onsite - System Design / Architecture

Apr 2026
System design round at Sesame for a software engineer role. The whole thing was one big question about building a voice chatbot from scratch, and it went pretty deep into areas I hadn't fully prepped for.
  • Design a voice chatbot end-to-end, covering speech-to-text, intent and dialog understanding, state management, response generation, text-to-speech, and audio streaming back to the user. Also address latency budgets per stage, barge-in and interruption handling, scaling for concurrent calls, and observability.

“This is one of those questions where you think you know the space until you actually have to draw it out.”

View Post

Sesame·Machine Learning Engineer·Take-home Assignment

Mar 2026
Sesame AI gave me a 90-minute take-home to build a DataLoader from scratch in three progressive stages. The scope kept expanding and I was not fully prepared for how deep the discussion questions would go on top of the implementation.
  • Implement a DataLoader that reads all JSON files from a directory and yields records in fixed-size batches.
  • Extend the DataLoader to support a deterministic shuffle controlled by a seed, and add a resume_from(index) method so the loader can pick up mid-epoch after a checkpoint.
  • Add dataset mixing: given multiple source directories with target sampling proportions, interleave records so the long-run distribution matches those proportions.
  • Discuss how shuffle interacts with resume, how mixing handles uneven dataset sizes with and without replacement, and what the memory footprint looks like across all three parts.

“This part felt manageable.”

View Post