As a Software Engineer at Skild AI, you are at the forefront of building the foundation for general-purpose artificial intelligence. This role is not merely about writing code; it is about architecting the systems that allow AI models to interact with, understand, and navigate the physical world. You will work on cutting-edge problems in robotics and machine learning, translating complex theoretical research into scalable, high-performance software systems. Your impact will be felt across the entire product lifecycle. By developing robust infrastructure and efficient algorithms, you enable Skild AI to push the boundaries of what autonomous systems can achieve. This position is both highly technical and deeply collaborative, requiring you to work alongside researchers and engineers to solve challenges that have no existing roadmap. If you are driven by the opportunity to shape the future of embodied AI and thrive in an environment of rapid iteration, this role is the perfect intersection of your skills and the company’s ambitious mission.
Recruiter Screening
reportedInitial screening to ensure alignment on experience and role expectations.
What to demonstrate
- Initial screening to ensure alignment on experience and role expectations
- Depth in Python Programming
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 Assessment
reportedLive coding session conducted in Python to assess technical capabilities.
What to demonstrate
- Live coding session conducted in Python to assess technical capabilities
- Depth in Python Programming
How to prepare
- Answer aloud and timed: Describe the trade-offs between different loss functions in a training pipeline.
- Answer aloud and timed: How would you handle data synchronization issues in a multi-modal robotic system?
Interview Loop
reportedComprehensive interviews including technical discussions, coding assessments, and behavioral evaluations.
What to demonstrate
- Comprehensive interviews including technical discussions, coding assessments, and behavioral evaluations
- Depth in Python Programming
How to prepare
- Answer aloud and timed: Explain how you manage memory and compute resources when training large-scale models.
- Answer aloud and timed: Implement a custom data loader for high-dimensional sensor data.
PracHub editorial advice for the preparation topics above.
Prioritize Clarity
When solving technical problems, always state your assumptions. This shows you are systematic and prevents you from going down the wrong path.
Focus on the "Why
Don't just show that you can code; explain why you chose a specific data structure or model architecture over others.
Stay Current
Keep up with the latest research in embodied AI, as it demonstrates genuine passion for the work Skild AI is doing.
Manage Your Energy
The final interview loop is demanding. Ensure you are well-rested and prepared for a full day of technical and behavioral assessment.
Choose a category, try a prompt, then open its approach, worked solution or follow-up when you need it.
Explain how you manage memory and compute resources when training large-scale models.
Explain how you manage memory and compute resources when training large-scale models.
Approach
- Say what the runtime actually does before reasoning about the code.
- Name what is shared across threads and what owns each piece of state.
- Identify the window where an invariant is briefly untrue.
- 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 a custom data loader for high-dimensional sensor data.
Implement a custom data loader for high-dimensional sensor data.
Approach
- Restate the input: its shape, its size, and what is guaranteed about it.
- Name the brute-force solution and its complexity before improving on it.
- Choose the data structure from the access pattern, not from familiarity.
- 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 process and filter noisy input streams from robotic sensors.
Write a function to process and filter noisy input streams from robotic sensors.
Approach
- Restate the input: its shape, its size, and what is guaranteed about it.
- Name the brute-force solution and its complexity before improving on it.
- Choose the data structure from the access pattern, not from familiarity.
- 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 structure a modular codebase for a complex AI research project?
How would you structure a modular codebase for a complex AI research project?
Approach
- Restate the input: its shape, its size, and what is guaranteed about it.
- Name the brute-force solution and its complexity before improving on it.
- Choose the data structure from the access pattern, not from familiarity.
- 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?
Solve a classic algorithmic problem, focusing on time and space complexity optimization.
Solve a classic algorithmic problem, focusing on time and space complexity optimization.
Approach
- Restate the input: its shape, its size, and what is guaranteed about it.
- Name the brute-force solution and its complexity before improving on it.
- Choose the data structure from the access pattern, not from familiarity.
- 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?
Refactor a provided snippet of code to improve readability and performance.
Refactor a provided snippet of code to improve readability and performance.
Approach
- Restate the input: its shape, its size, and what is guaranteed about it.
- Name the brute-force solution and its complexity before improving on it.
- Choose the data structure from the access pattern, not from familiarity.
- 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
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
- 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.
- 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.
- 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.
- 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?
Write the update path that detects a concurrent edit
resource carries version INT NOT NULL DEFAULT 1. resource_revision holds revision_id, resource_id, version, actor_user_id, change_kind, patch JSONB, request_id, created_at with UNIQUE (resource_id, version). outbox_event holds aggregate_type, aggregate_id, aggregate_version, event_type, payload, status. A PUT carries the version the client read. Write the exact statements for the single transaction that applies the edit, records the revision and enqueues 'resource.updated', and give the handler's branch on zero affected rows. Then say what PostgreSQL 16 does under READ COMMITTED when two of these updates hit one row at once.
Approach
- One transaction, three writes, no network call inside it: UPDATE resource SET title = $3, version = version + 1, updated_at = now() WHERE resource_id = $1 AND tenant_id = $4 AND version = $2; then INSERT the resource_revision row at version $2 + 1; then INSERT the outbox_event row at the same aggregate_version. The event goes to a table rather than a broker because no transaction spans both.
- Branch on the affected-row count before doing anything else. Zero has three causes — stale version, wrong tenant, row gone — so re-read once and map to 409 carrying the current version, or 404 for an id outside the caller's tenant, which also stops the endpoint confirming that another tenant's id exists.
- State the engine behaviour instead of assuming it. Under READ COMMITTED the second UPDATE blocks on the row lock, and when the first commits PostgreSQL re-evaluates the WHERE clause against the newly committed row, so the version predicate now fails and the statement reports zero rows. Under REPEATABLE READ the identical collision raises SQLSTATE 40001 instead, so the handler must fold both shapes into one conflict response.
- Keep UNIQUE (resource_id, version) even though the predicate already serialises writers. It is what makes a lost update unwritable if any other path ever reaches the revision table, and it converts a logic bug into 23505 rather than into a silently missing history row.
Follow-up
- A client sends the version it read ten minutes ago and the resource has moved three versions. What is in your 409 so it can resolve the conflict without a full re-fetch?
- Two editors, two disjoint fields, no overlap. Does your answer still refuse the second write, and should it?
Explain the architecture of a transformer and how it applies to robotic control.
Explain the architecture of a transformer and how it applies to robotic control.
Approach
- Fix the scope first: who calls this, how often, and what they do when it fails.
- Name the read and write paths separately; they rarely have the same bottleneck.
- Choose a partition key and say what query it makes expensive.
- 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 optimize inference latency in a production environment?
How do you optimize inference latency in a production environment?
Approach
- Clarify what is being asked and what a complete answer contains.
- State your assumptions explicitly before working the problem.
- Say what you would check first and why it is the highest-information step.
- Work from the requirement backwards to the design.
Follow-up
- What assumption would you test first?
- How would you know your answer was wrong?
Describe the trade-offs between different loss functions in a training pipeline.
Describe the trade-offs between different loss functions in a training pipeline.
Approach
- Clarify what is being asked and what a complete answer contains.
- State your assumptions explicitly before working the problem.
- Say what you would check first and why it is the highest-information step.
- 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 would you handle data synchronization issues in a multi-modal robotic system?
How would you handle data synchronization issues in a multi-modal robotic system?
Approach
- Clarify what is being asked and what a complete answer contains.
- State your assumptions explicitly before working the problem.
- Say what you would check first and why it is the highest-information step.
- Work from the requirement backwards to the design.
Follow-up
- What assumption would you test first?
- How would you know your answer was wrong?
Listing latency scales with page size, not with filters
The tenant listing endpoint reads resource filtered by tenant_id and status, ordered by updated_at DESC, and returns each row plus the owner's display name from app_user and the actor of that resource's latest resource_revision. p99 is 55 ms at 10 rows per page and 1.4 s at 200. Database telemetry shows 401 statements per request, each under 1 ms, and nothing in the slow-query log. Diagnose the cause and give the fix, stating the statement count per request and the p99 you expect afterwards.
Approach
- Read the counters before forming a theory. 401 statements for 200 rows is one driver query plus two per row, and sub-millisecond execution with an empty slow-query log rules out a bad plan. The time is round trips, which is why it is invisible in every per-query metric and scales with rows returned rather than with filter selectivity.
- Name the two per-row statements from their normalised text: a single-row app_user lookup by user_id, and a resource_revision lookup by resource_id ordered by version DESC LIMIT 1. Confirm by dropping those two response fields and watching the statement count fall to one. That locates the calls in the serialisation layer, not the repository.
- Check that the arithmetic accounts for the whole gap. Measure one round trip to the replica in isolation; 400 trips at roughly 3 ms of network plus 0.2 ms of execution is about 1.3 s on top of a 55 ms baseline, which matches. If the multiplication had fallen short, the N+1 would only be part of the story and you would keep looking.
- Batch both lookups. Collect owner_user_ids and resource_ids from the driver query, then issue WHERE tenant_id = $1 AND user_id = ANY($2) for the users, and PostgreSQL's SELECT DISTINCT ON (resource_id) ... WHERE resource_id = ANY($2) ORDER BY resource_id, version DESC for the latest revision, which the UNIQUE (resource_id, version) index serves directly. On an engine without DISTINCT ON, use a lateral join or a row_number window. Three statements per request at any page size.
Follow-up
- The page size is capped at 200 today. What breaks first if it is raised to 2,000, and is it still this bug?
- How do you stop the next N+1 from reaching production, given that no individual query is slow and the endpoint's tests pass?
Built from the rounds and topics Skild AI candidates report.
Prepare, practise & reflect
One practical outcome each day. Spend longer where you need it.
0 / 7 done01Map the Skild AI loop
- Write out the reported sequence: Recruiter Screening, Technical Assessment, Interview 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.
02Work Python Programming
- Spend the session on Python Programming, which Skild AI candidates report being tested on.
- Write one worked example in Python Programming and time yourself on it.
Deliverable: One timed worked example in Python Programming.
03Work Live Coding Interviews
- Spend the session on Live Coding Interviews, which Skild AI candidates report being tested on.
- Write one worked example in Live Coding Interviews and time yourself on it.
Deliverable: One timed worked example in Live Coding Interviews.
04Work Coding Interviews (General)
- Spend the session on Coding Interviews (General), which Skild AI candidates report being tested on.
- Write one worked example in Coding Interviews (General) and time yourself on it.
Deliverable: One timed worked example in Coding Interviews (General).
05Answer out loud: Technical and Domain Knowledge
- Answer aloud, timed: Explain the architecture of a transformer and how it applies to robotic control.
- Answer aloud, timed: How do you optimize inference latency in a production environment?
Deliverable: Spoken answers to 2 reported Technical and Domain Knowledge question(s), under time.
06Answer out loud: Coding and Algorithms
- Answer aloud, timed: Implement a custom data loader for high-dimensional sensor data.
- Answer aloud, timed: Write a function to process and filter noisy input streams from robotic sensors.
Deliverable: Spoken answers to 2 reported Coding and Algorithms question(s), under time.
07Dry run for Skild AI
- Run one full mock under time, then write down the two questions you most want to ask your interviewers.
Deliverable: A completed timed mock and two questions to ask.
Expand any day for tasks and deliverables. Your progress is saved on this device.
Behavioural rounds judge the decision you made and what it cost.
Turn a code review disagreement into a decision
A colleague's change updates a row with UPDATE resource SET version = version + 1 WHERE resource_id = $1 AND version = $2 and treats an affected-row count of zero as a successful no-op. You read that as a silently lost update; they think returning 200 is friendlier to clients than returning a conflict. Describe how you have handled a review disagreement of this shape: what goes in the comment, when you leave the thread, and who decides. Then write the comment you would leave here, in under 80 words.
Approach
- Sort the disagreement before writing anything. A silently discarded write is a correctness claim about data; the choice between 409 and 412 is taste. Only the first justifies blocking a merge, and saying which one you are doing is most of the value of the comment.
- Make the claim reproducible in the comment itself with an interleaving rather than a principle: A reads version 7, B reads version 7, B commits version 8, A's predicate matches zero rows, A is told it succeeded and A's edit is gone.
- Offer the alternative with its cost attached: return 409 carrying the current version and the revision that won, so the client can re-read and re-apply. Note that automatic retry is not the fix, because a retry re-reads the winner's state and reapplies an intent formed against data that no longer exists.
- Apply an escalation rule you can state: two round trips on the thread, then a call, and the service's owner decides rather than the reviewer. A reviewer who cannot be overruled is a bottleneck with extra steps.
Follow-up
- Where would you put the test that fails if someone reintroduces the swallowed zero rowcount?
- The author says clients cannot handle a 409. How do you check whether that is true?
Narrate an outage you owned from page to postmortem
Pick an incident you personally drove, ideally one where writes were affected rather than reads. In six to eight minutes: state the symptom as it first appeared on a dashboard, the blast radius you established before you knew the cause, the mitigation you applied and when, the mechanism you eventually proved, and the follow-up that would prevent a repeat. Bring numbers: error rate, tenants affected, minutes to mitigate, minutes to resolve. If you cannot name what you measured, choose a different incident.
Approach
- Open on the signal rather than the cause: which metric at which percentile moved, on which service, at what time, so the listener follows the same evidence you had rather than a conclusion you already reached.
- Separate mitigation from diagnosis out loud. State what you did to stop the bleeding (flag off, shed traffic, drain a lease, roll back a deploy) and say plainly that you did it before the mechanism was known, because those are two jobs with different deadlines.
- Establish blast radius in countable terms: how many tenants, how many writes, and crucially whether the effect was loss or only delay. An append-only revision table or a pending outbox row means the change survived and the projection was merely behind, which is a repair rather than a data-loss incident.
- Prove the mechanism instead of asserting it. Name the trace span that grew, the plan that flipped to a sequential scan, the lease that expired, plus one alternative you ruled out and the signal that stayed flat while you ruled it out.
Follow-up
- What would you do differently in the first five minutes, given the same dashboard and no more information?
- Which follow-up action did you deliberately not take, and why was dropping it the right call?
Ship under a deadline and bound the debt you chose
You have four days to ship a tenant-facing listing endpoint. The version you would defend uses keyset pagination over (tenant_id, status, updated_at DESC, resource_id DESC); the version you can finish uses LIMIT/OFFSET with no matching index. Describe a deadline call you actually made of this shape: what you shipped, what you knowingly deferred, how you bounded the damage with a mechanism rather than an intention, and the specific numeric condition that would force the follow-up. Name who you told and where you wrote it down.
Approach
- Name the deferred failure precisely instead of calling it slow. OFFSET n makes the database produce and discard n rows, so cost grows with page depth; without an index matching the sort, every matching row is read and sorted before the limit applies; and rows inserted between two page fetches shift across the boundary so items are skipped or repeated with nothing in the response to signal it.
- Bound the blast radius with something mechanical rather than a promise: cap maximum page depth, cap page size, restrict the endpoint to one internal caller, or keep it behind a flag. State which failure each cap removes and which it leaves standing.
- Attach a number to the trigger and wire it to an alarm: the first tenant crossing N resources, or the endpoint's p99 crossing its share of the 400 ms budget, so the debt announces itself instead of waiting to be remembered.
- Write it where the next engineer looks, which is the code and the ticket, not a chat message: what was deferred, why, the cap, and the trigger.
Follow-up
- At what page depth does the offset version breach your latency budget, given your page size and row counts?
- What breaks first when you switch to keyset pagination later, and what does a client holding an old page token see?
- 01
A colleague's change updates a row with UPDATE resource SET version = version + 1 WHERE resource_id = $1 AND version = $2 and treats an affected-row count of zero as a successful no-op. You read that as a silently lost update; they think returning 200 is friendlier to clients than returning a conflict. Describe how you have handled a review disagreement of this shape: what goes in the comment, when you leave the thread, and who decides. Then write the comment you would leave here, in under 80 words.
- 02
Pick an incident you personally drove, ideally one where writes were affected rather than reads. In six to eight minutes: state the symptom as it first appeared on a dashboard, the blast radius you established before you knew the cause, the mitigation you applied and when, the mechanism you eventually proved, and the follow-up that would prevent a repeat. Bring numbers: error rate, tenants affected, minutes to mitigate, minutes to resolve. If you cannot name what you measured, choose a different incident.
- 03
You have four days to ship a tenant-facing listing endpoint. The version you would defend uses keyset pagination over (tenant_id, status, updated_at DESC, resource_id DESC); the version you can finish uses LIMIT/OFFSET with no matching index. Describe a deadline call you actually made of this shape: what you shipped, what you knowingly deferred, how you bounded the damage with a mechanism rather than an intention, and the specific numeric condition that would force the follow-up. Name who you told and where you wrote it down.
How long is the interview process?
The process can vary, but generally expect it to span several weeks from the initial screen to the final decision. Be prepared for potential fluctuations in timeline due to the company’s rapid growth and high interview volume.
Skild AI Software Engineer candidate reports ↗What is the best way to prepare for the coding rounds?
Focus on solving problems in Python that require data manipulation and algorithmic efficiency. Practice explaining your thought process out loud, as interviewers value your ability to communicate complex ideas clearly.
Skild AI Software Engineer candidate reports ↗Is there a specific focus on robotics?
While the core is Software Engineering, the application is robotics. Familiarity with the challenges of real-world AI, such as latency, sensor data, and physical constraints, will significantly differentiate your candidacy. ##### Tip The interview process can be intensive and sometimes lengthy; stay proactive in your communication with the recruiter to manage expectations throughout the pipeline.
Skild AI Software Engineer candidate reports ↗What topics does Skild AI test in interviews?
Skild AI interviews most often cover Reinforcement Learning (RL), Python Programming, Robotics Machine Learning, Live Coding Interviews, and Training Foundation Models at Scale. The exact emphasis depends on the specific role you apply for.
Skild AI Software Engineer candidate reports ↗Sources & methodology 3 sources ↗
No official company page is cited. Rounds and questions come from candidate reports and PracHub editorial material; each source shows the date it was read.
- 01Skild AI Software Engineer candidate reports ↗
Company-reported rounds, questions and FAQ.
Candidate reports · Accessed 2026-09-22 - 02PracHub Software Engineer practice ↗
PracHub practice material, not company-reported.
PracHub page · Accessed 2026-09-22 - 03PracHub preparation framework ↗
PracHub preparation guidance.
PracHub page · Accessed 2026-09-22
Explore related engineering environments and compare the preparation focus.
Scale Ai Software Engineer Interview Guide
Data structures and algorithms practice for Scale Ai candidates, where stating a bound means stating its precondition: constant-time hash lookup is an average case over keys that spread.
Lambda Software Engineer Interview Guide
Failure-handling practice for Lambda software engineer candidates: requests whose outcome a timeout leaves unknown, where a retry is safe, and deduplicating at the receiver instead of assuming exactly-once delivery.
Discord Software Engineer Interview Guide
Database practice for Discord candidates: how an index changes a query plan, and which anomalies survive a given isolation level, since read committed still permits non-repeatable reads.
CoreWeave Software Engineer Interview Guide
Failure-handling practice for CoreWeave software engineer candidates: requests whose outcome a timeout leaves unknown, where a retry is safe, and deduplicating at the receiver instead of assuming exactly-once delivery.