cswork / libraryOpen practice workspace

Interview guides · Zyphra

Zyphra Software Engineer Interview Guide

Zyphra Software Engineer interview guide: reported rounds, questions and how candidates describe the process.

The 60-second brief

At Zyphra, a Software Engineer is at the absolute center of the next generation of artificial intelligence. We are building state-of-the-art AI models, high-performance computing platforms, and developer-facing interfaces that make advanced intelligence accessible and scalable. Our engineering team does not just consume existing technologies; we design the underlying systems, platforms, and interfaces that define how enterprises and developers interact with cutting-edge AI. Whether you are optimizing low-level performance on GPU clusters, building highly responsive full-stack interfaces, or designing APIs that support millions of requests, your work has a direct impact on the speed of AI adoption. The systems we build must handle massive scale, complex data pipelines, and real-time model inference.

This guide is scoped to a Software Engineer candidate at Zyphra.

Focus areas

  • Full-Stack Development
  • Frontend Engineering
  • Platform Engineering

The Zyphra interview process

Zyphra candidates report 3 rounds over 3-5 weeks. The stages below are what candidates describe, not a published process.

Initial Screen

Technical and behavioral screen with a recruiter or hiring manager to align on your background and interest.

What to demonstrate

  • Technical and behavioral screen with a recruiter or hiring manager to align on your background and interest
  • Depth in Full-Stack Development

How to prepare

  • Be able to walk your CV end to end in two minutes, and say why this company specifically.
  • Have your salary expectations, notice period and location constraints ready, and ask for the rest of the loop in writing.

Technical Assessments

Complete one or two deep-dive technical assessments, which may include live coding, system design, or a practical take-home project.

What to demonstrate

  • Complete one or two deep-dive technical assessments, which may include live coding, system design, or a practical take-home project
  • Depth in Full-Stack Development

How to prepare

  • Answer aloud and timed: How would you optimize a data ingestion pipeline that processes terabytes of unstructured data for model training?
  • Answer aloud and timed: Describe a scenario where you had to debug a memory leak or bottleneck in a high-throughput production system.

Onsite/Virtual Loop

Comprehensive loop where you meet with cross-functional team members, discuss architecture, and assess cultural alignment.

What to demonstrate

  • Comprehensive loop where you meet with cross-functional team members, discuss architecture, and assess cultural alignment
  • Depth in Full-Stack Development

How to prepare

  • Answer aloud and timed: How do you manage state and consistency across a distributed microservices architecture?
  • Answer aloud and timed: How would you structure the frontend state management for a web application that displays real-time, streaming outputs from an AI model?

Where candidates lose points

Over-communicate your thought process

During coding and system design interviews, talk through your ideas, trade-offs, and assumptions before writing any code. We want to see how you approach problems, not just your final solution.

Focus on trade-offs

There is rarely a single "correct" answer in system design. Explain why you chose a specific database, network protocol, or architectural pattern, and discuss the alternatives you rejected.

Show product and developer empathy

When designing interfaces or APIs, keep the end-user in mind. Explain how your design choices make the platform more intuitive, reliable, or performant for developers.

Questions & practice

Explain the trade-offs between synchronous and asynchronous processing when deploying a large-scale machine le

Medium · Systems & Platform Engineering

Explain the trade-offs between synchronous and asynchronous processing when deploying a large-scale machine learning model for inference.

Approach

  1. Say what the runtime actually does before reasoning about the code.
  2. Name what is shared across threads and what owns each piece of state.
  3. Identify the window where an invariant is briefly untrue.
  4. Distinguish a value from a reference to it, and say which one you handed out.

Follow-up

  • What happens if two callers reach this at the same time?
  • Where could this allocate more than you expect?

Implement an algorithm to find the shortest path in a weighted graph, and discuss its time and space complexit

Medium · Problem-Solving & Algorithms

Implement an algorithm to find the shortest path in a weighted graph, and discuss its time and space complexity.

Approach

  1. Restate the input: its shape, its size, and what is guaranteed about it.
  2. Name the brute-force solution and its complexity before improving on it.
  3. Choose the data structure from the access pattern, not from familiarity.
  4. State the target complexity and say which constraint rules the naive version out.

Follow-up

  • How does this change if the input no longer fits in memory?
  • What is the worst case, and how likely is it on real data?

Write a function to serialize and deserialize a binary tree, ensuring optimal space usage.

Medium · Problem-Solving & Algorithms

Approach

  1. Restate the input: its shape, its size, and what is guaranteed about it.
  2. Name the brute-force solution and its complexity before improving on it.
  3. Choose the data structure from the access pattern, not from familiarity.
  4. State the target complexity and say which constraint rules the naive version out.

Follow-up

  • How does this change if the input no longer fits in memory?
  • What is the worst case, and how likely is it on real data?

How would you optimize a search algorithm to run in parallel across multiple CPU cores?

Medium · Problem-Solving & Algorithms

Approach

  1. Restate the input: its shape, its size, and what is guaranteed about it.
  2. Name the brute-force solution and its complexity before improving on it.
  3. Choose the data structure from the access pattern, not from familiarity.
  4. State the target complexity and say which constraint rules the naive version out.

Follow-up

  • How does this change if the input no longer fits in memory?
  • What is the worst case, and how likely is it on real data?

Denormalise tenant onto revisions and backfill it live

Hard · denormalisation · online migration · backfill

resource_revision (revision_id, resource_id, version, actor_user_id, change_kind, patch, request_id, created_at) has 400M rows and no tenant column; tenant_id lives only on resource. Two reads need it: a tenant-scoped audit feed ordered by created_at DESC, and an offboarding purge. Both join back to resource today. Justify adding tenant_id to resource_revision against those two reads, name the anomaly the copy introduces and the constraint that prevents it, then give the ordered migration for a live table taking 1.2k writes/second — the lock each step takes, how the backfill is batched, and where each step stops being reversible. PostgreSQL 16.

Approach

  1. Justify from the access path rather than from taste. Without the column, the audit feed either scans resource_revision by created_at and discards other tenants' rows, or resolves the tenant's resource_ids first and probes with them — both proportional to the tenant's whole history rather than to one page. With (tenant_id, created_at DESC, revision_id DESC) it is a seek that stops at 50 rows, and the purge becomes a ranged delete instead of a join.
  2. Name the cost exactly: a second copy of a fact can disagree with the first. Make the disagreement unwritable rather than documented — add UNIQUE (resource_id, tenant_id) on resource so it can serve as a foreign-key target, then FOREIGN KEY (resource_id, tenant_id) REFERENCES resource (resource_id, tenant_id) on the revision table. A revision can then only ever carry its parent's tenant.
  3. Step one, expand: ALTER TABLE resource_revision ADD COLUMN tenant_id BIGINT NULL, with no default, so it is a catalogue change and no rewrite. It still needs ACCESS EXCLUSIVE for an instant, and that instant queues behind the longest open transaction on the table while every later query queues behind it — set lock_timeout to 2s and retry rather than wait.
  4. Step two, dual-write: deploy the writer that populates tenant_id on every new revision while reads still use the join. Reversible by redeploying the previous build, because nothing reads the column yet.

Follow-up

  • The backfill is half finished and a rollback is required. What state is the table in, and what does the previous build do with a half-populated column?
  • How do you verify the backfill actually finished, given rows are still being inserted while it runs?

Stop tag and share joins from fanning out a page

Medium · join fan-out · semi-joins · many-to-many

resource_tag is (resource_id, tag_id) with PK (resource_id, tag_id); resource_share is (resource_id, shared_with_user_id, permission). The tagged-and-shared listing inner-joins resource to both, filters tenant_id, tag_id = ANY($2) and shared_with_user_id = $3, orders by updated_at DESC and takes 50. Pages come back with fewer than 50 distinct resources and the total in the header is far too high. Explain the row multiplication, rewrite both the page query and the count query so each is correct, and name the index each one needs. PostgreSQL 16.

Approach

  1. Do the arithmetic against the predicates that are actually there. An inner join emits one row per matching child row, and both joins are filtered: tag_id = ANY($2) admits only the requested tags, shared_with_user_id = $3 admits one user's share rows. So a resource holding three of the requested tags and shared with $3 once yields three rows, not one — the multiplier is its count of matching tags times its share rows for that single user, and that second factor is 1 unless the table admits duplicate (resource_id, shared_with_user_id) pairs. LIMIT 50 then limits rows rather than resources, and COUNT(*) counts pairs — the header is the product, not the population.
  2. Reject DISTINCT as the fix. It deduplicates after the product has been built, so the planner must materialise and sort the fanned-out set before the LIMIT can apply, and it leaves any SUM or AVG in the same select list wrong.
  3. Rewrite both filters as semi-joins, keeping resource as the only row source: AND EXISTS (SELECT 1 FROM resource_tag rt WHERE rt.resource_id = r.resource_id AND rt.tag_id = ANY($2)) and the same shape against resource_share. A semi-join stops at the first match per resource and preserves the driving index order, so ORDER BY updated_at DESC, resource_id DESC LIMIT 50 still stops after 50 rows.
  4. Count with the same predicates and no join at all: SELECT count(*) FROM resource r WHERE r.tenant_id = $1 AND r.status = 'active' AND EXISTS (...) AND EXISTS (...). Nothing multiplies a resource, so the number is the population.

Follow-up

  • The filter changes from 'any of these tags' to 'all of these tags'. Rewrite it and state what it costs relative to the ANY form.
  • A resource can be shared with the same user twice under different permissions. Does your count change, and should it?

How would you design a distributed rate limiter that can handle tens of thousands of requests per second acros

Medium · Systems & Platform Engineering

How would you design a distributed rate limiter that can handle tens of thousands of requests per second across multiple regions?

Approach

  1. Fix the scope first: who calls this, how often, and what they do when it fails.
  2. Name the read and write paths separately; they rarely have the same bottleneck.
  3. Choose a partition key and say what query it makes expensive.
  4. State the consistency you need, and where you are willing to be stale.

Follow-up

  • What breaks first when traffic grows ten times?
  • How does this behave when that dependency is down for an hour?

How would you optimize a data ingestion pipeline that processes terabytes of unstructured data for model train

Medium · Systems & Platform Engineering

How would you optimize a data ingestion pipeline that processes terabytes of unstructured data for model training?

Approach

  1. Clarify what is being asked and what a complete answer contains.
  2. State your assumptions explicitly before working the problem.
  3. Say what you would check first and why it is the highest-information step.
  4. Work from the requirement backwards to the design.

Follow-up

  • What assumption would you test first?
  • How would you know your answer was wrong?

How do you manage state and consistency across a distributed microservices architecture?

Medium · Systems & Platform Engineering

Approach

  1. Fix the scope first: who calls this, how often, and what they do when it fails.
  2. Name the read and write paths separately; they rarely have the same bottleneck.
  3. Choose a partition key and say what query it makes expensive.
  4. State the consistency you need, and where you are willing to be stale.

Follow-up

  • What breaks first when traffic grows ten times?
  • How does this behave when that dependency is down for an hour?

How would you structure the frontend state management for a web application that displays real-time, streaming

Medium · Frontend & Full-Stack Architecture

How would you structure the frontend state management for a web application that displays real-time, streaming outputs from an AI model?

Approach

  1. Fix the scope first: who calls this, how often, and what they do when it fails.
  2. Name the read and write paths separately; they rarely have the same bottleneck.
  3. Choose a partition key and say what query it makes expensive.
  4. State the consistency you need, and where you are willing to be stale.

Follow-up

  • What breaks first when traffic grows ten times?
  • How does this behave when that dependency is down for an hour?

Design an API layer that supports both REST and WebSockets, explaining how you would handle connection dropout

Medium · Frontend & Full-Stack Architecture

Design an API layer that supports both REST and WebSockets, explaining how you would handle connection dropouts and retries.

Approach

  1. Say who the caller is and what they do when the call fails halfway.
  2. Define the identity of a request so a retry cannot double-apply it.
  3. Separate accepted, pending, failed and confirmed; they are different facts.
  4. Design the error taxonomy before the success shape; callers branch on it.

Follow-up

  • What happens if the caller retries after a timeout?
  • How does a client discover it is on an old version of this contract?

What strategies do you use to optimize the rendering performance of complex data visualizations or large inter

Medium · Frontend & Full-Stack Architecture

What strategies do you use to optimize the rendering performance of complex data visualizations or large interactive dashboards?

Approach

  1. Fix the scope first: who calls this, how often, and what they do when it fails.
  2. Name the read and write paths separately; they rarely have the same bottleneck.
  3. Choose a partition key and say what query it makes expensive.
  4. State the consistency you need, and where you are willing to be stale.

Follow-up

  • What breaks first when traffic grows ten times?
  • How does this behave when that dependency is down for an hour?

How do you ensure security and prevent common vulnerabilities (like XSS or CSRF) when building full-stack appl

Medium · Frontend & Full-Stack Architecture

How do you ensure security and prevent common vulnerabilities (like XSS or CSRF) when building full-stack applications?

Approach

  1. Fix the scope first: who calls this, how often, and what they do when it fails.
  2. Name the read and write paths separately; they rarely have the same bottleneck.
  3. Choose a partition key and say what query it makes expensive.
  4. State the consistency you need, and where you are willing to be stale.

Follow-up

  • What breaks first when traffic grows ten times?
  • How does this behave when that dependency is down for an hour?

Describe how you would build a reusable component library that supports multiple themes and strict accessibili

Medium · Frontend & Full-Stack Architecture

Describe how you would build a reusable component library that supports multiple themes and strict accessibility standards.

Approach

  1. Fix the scope first: who calls this, how often, and what they do when it fails.
  2. Name the read and write paths separately; they rarely have the same bottleneck.
  3. Choose a partition key and say what query it makes expensive.
  4. State the consistency you need, and where you are willing to be stale.

Follow-up

  • What breaks first when traffic grows ten times?
  • How does this behave when that dependency is down for an hour?

Given a stream of data, how would you design a system to find the top K most frequent elements in real time?

Medium · Problem-Solving & Algorithms

Approach

  1. Fix the scope first: who calls this, how often, and what they do when it fails.
  2. Name the read and write paths separately; they rarely have the same bottleneck.
  3. Choose a partition key and say what query it makes expensive.
  4. State the consistency you need, and where you are willing to be stale.

Follow-up

  • What breaks first when traffic grows ten times?
  • How does this behave when that dependency is down for an hour?

Describe a scenario where you had to debug a memory leak or bottleneck in a high-throughput production system.

Medium · Systems & Platform Engineering

Approach

  1. Establish what changed and when, before forming any theory.
  2. Pick a bisection that eliminates candidates whichever way it turns out.
  3. Check the instrumentation before believing the symptom.
  4. Separate the trigger from the cause; the deploy is rarely the bug.

Follow-up

  • What would you look at first, and what would it rule out?
  • How would you tell a cause from a coincidence here?

Suggested practice checklist

Built from the rounds and topics Zyphra candidates report.

Day 1 — Map the Zyphra loop

  • Write out the reported sequence: Initial Screen, Technical Assessments, Onsite/Virtual Loop.
  • For each round, write one sentence on what it is judging, from the description above, and mark the one you are least ready for.

Deliverable: A one-page map of the 3 reported rounds, with the weakest marked.

Day 2 — Work Full-Stack Development

  • Spend the session on Full-Stack Development, which Zyphra candidates report being tested on.
  • Write one worked example in Full-Stack Development and time yourself on it.

Deliverable: One timed worked example in Full-Stack Development.

Day 3 — Work Frontend Engineering

  • Spend the session on Frontend Engineering, which Zyphra candidates report being tested on.
  • Write one worked example in Frontend Engineering and time yourself on it.

Deliverable: One timed worked example in Frontend Engineering.

Day 4 — Work Platform Engineering

  • Spend the session on Platform Engineering, which Zyphra candidates report being tested on.
  • Write one worked example in Platform Engineering and time yourself on it.

Deliverable: One timed worked example in Platform Engineering.

Day 5 — Answer out loud: Systems & Platform Engineering

  • Answer aloud, timed: How would you design a distributed rate limiter that can handle tens of thousands of requests per second across multiple regions?
  • Answer aloud, timed: Explain the trade-offs between synchronous and asynchronous processing when deploying a large-scale machine learning model for inference.

Deliverable: Spoken answers to 2 reported Systems & Platform Engineering question(s), under time.

Day 6 — Answer out loud: Frontend & Full-Stack Architecture

  • Answer aloud, timed: How would you structure the frontend state management for a web application that displays real-time, streaming outputs from an AI model?
  • Answer aloud, timed: Design an API layer that supports both REST and WebSockets, explaining how you would handle connection dropouts and retries.

Deliverable: Spoken answers to 2 reported Frontend & Full-Stack Architecture question(s), under time.

Day 7 — Answer out loud: Problem-Solving & Algorithms

  • Answer aloud, timed: Implement an algorithm to find the shortest path in a weighted graph, and discuss its time and space complexity.
  • Answer aloud, timed: Given a stream of data, how would you design a system to find the top K most frequent elements in real time?

Deliverable: Spoken answers to 2 reported Problem-Solving & Algorithms question(s), under time.

Behavioural preparation

Behavioural rounds judge the decision you made and what it cost.

Describe a time when you had to make a critical technical decision with incomplete information or shifting req

Medium · Behavioral & Alignment

Describe a time when you had to make a critical technical decision with incomplete information or shifting requirements.

Approach

  1. Pick a story where you made the decision, not one where you watched it.
  2. State the situation in two sentences and spend the rest on the reasoning.
  3. Give the blast radius: what could have broken, and what you measured.
  4. Name the disagreement and how you resolved it with evidence.

Follow-up

  • What would you do differently if you ran that again?
  • How did you know your change caused the improvement?

How do you handle disagreements with product managers or researchers regarding technical feasibility versus fe

Medium · Behavioral & Alignment

How do you handle disagreements with product managers or researchers regarding technical feasibility versus feature scope?

Approach

  1. Pick a story where you made the decision, not one where you watched it.
  2. State the situation in two sentences and spend the rest on the reasoning.
  3. Give the blast radius: what could have broken, and what you measured.
  4. Name the disagreement and how you resolved it with evidence.

Follow-up

  • What would you do differently if you ran that again?
  • How did you know your change caused the improvement?

Tell me about a complex technical concept you had to explain to a non-technical stakeholder or client.

Medium · Behavioral & Alignment

Approach

  1. Pick a story where you made the decision, not one where you watched it.
  2. State the situation in two sentences and spend the rest on the reasoning.
  3. Give the blast radius: what could have broken, and what you measured.
  4. Name the disagreement and how you resolved it with evidence.

Follow-up

  • What would you do differently if you ran that again?
  • How did you know your change caused the improvement?

What is your approach to mentoring junior engineers while maintaining a high personal output?

Medium · Behavioral & Alignment

Approach

  1. Pick a story where you made the decision, not one where you watched it.
  2. State the situation in two sentences and spend the rest on the reasoning.
  3. Give the blast radius: what could have broken, and what you measured.
  4. Name the disagreement and how you resolved it with evidence.

Follow-up

  • What would you do differently if you ran that again?
  • How did you know your change caused the improvement?