Katkat

Company

Together AI interview questions

9 practice questions in the style of Together AI's interviews, covering System Design, Coding & Algorithms, Software Engineering Fundamentals.

Questions9

5 easy3 medium1 hard

Your progress0/9

Together AI interview questions by the numbers

Last updated

Katkat has 9 Together AI interview questions. The largest topics are Coding & Algorithms (44%), Software Engineering Fundamentals (44%) and System Design (11%). By difficulty: 56% easy, 33% medium and 11% hard. The most common round is Technical Screen (100%). Frequent concepts include Greedy, Hash Table, Simulation and Sorting.

Topics

Difficulty

  • Easy56% · 5
  • Medium33% · 3
  • Hard11% · 1

Interview rounds

  • Technical Screen100% · 9

Roles

  • Software Engineer100% · 9

Frequent concepts

  • Greedy11% · 1
  • Hash Table11% · 1
  • Simulation11% · 1
  • Sorting11% · 1
  • System Design11% · 1

Showing 1–9 of 9 results

  • Design a GPU-aware pod schedulerHard

    This interview question evaluates requirements, scale assumptions, API/data design, architecture, trade-offs, failure modes, and rollout in a realistic interview setting. A strong answer for Design a GPU-aware pod scheduler states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

    Together AISoftware EngineerSystem DesignTechnical Screen
  • Detect cycles and break them in pod dependenciesMedium

    This question evaluates proficiency with directed graph concepts and dependency resolution, focusing on cycle detection and selecting a single dependency edge whose removal renders the graph acyclic.

    Together AISoftware EngineerCoding & AlgorithmsTechnical Screen
  • Diagnose CPU Performance Problems on a Linux ServerEasy

    Diagnose Linux CPU slowness from host and thread metrics through scheduling, limits, profiles, and a workload-linked verification of the bottleneck.

    Together AISoftware EngineerSoftware Engineering FundamentalsTechnical Screen
  • Diagnose Slow Network Throughput Including ContainersEasy

    Investigate slow network throughput across application, transport, host, and container layers using controlled comparisons, retransmissions, limits, and path behavior.

    Together AISoftware EngineerSoftware Engineering FundamentalsTechnical Screen
  • Diagnose Storage I/O Performance and Identify the WorkloadEasy

    Diagnose storage slowness using device metrics, process attribution, workload shape, cache and durability effects, and measurements appropriate to the storage stack.

    Together AISoftware EngineerSoftware Engineering FundamentalsTechnical Screen
  • Historical Lookups in a Timestamped Key-Value StoreMedium

    Implement historical key-value lookups using timestamped writes, per-key ordered histories, and binary search for the latest eligible value.

    Together AISoftware EngineerCoding & AlgorithmsTechnical Screen
  • Investigate a Server Saturated by LoggingEasy

    Trace logging-related server saturation through resource attribution, producer/sink behavior, bounded queues, rotation and retention, and verified mitigation.

    Together AISoftware EngineerSoftware Engineering FundamentalsTechnical Screen
  • Schedule GPU Pods and Drain a NodeMedium

    Implement a GPU cluster manager to schedule workloads using a deterministic best-fit policy and simulate draining a node to check if all its pods can be migrated without disrupting other nodes.

    Together AISoftware EngineerCoding & AlgorithmsTechnical Screen
  • Split a Chunked Text Stream into Line-Balanced PartsEasy

    Construct any balanced whole-line partition of a chunked text stream, preserving exact content and newline rules while explaining buffering and streaming limits.

    Together AISoftware EngineerCoding & AlgorithmsTechnical Screen

More companies