Python interviews

Work in the tool while you explain your decisions.

Choose a role and level. Each interview has a new task, follow-up questions and a report that checks your work and your explanation.

Take the assessment Choose an interview

Speak or type. Sign in to save your results across devices.

Timed assessment

12 practical questions
50 minutes
Mid level

Questions are drawn from a bank of 36. Your report includes a skill breakdown and lessons to practice next.

Start assessment

Practice interviews

Junior tasks cover the fundamentals. Mid and senior tasks add debugging, changing requirements and review decisions.

Position

10 interviews

Software engineer
Junior
20 min

Stable participant identifiers

Clover Events needs participant_ids(values). Return unique nonempty identifiers in first-seen order after trimming and casefolding strings. Ignore None. Don't mutate the input. Repeated spellings identify one participant.

Start interview

Skills and rubricOrder-preserving identifier normalization

Keep a separate membership set and ordered result. Normalize before checking membership.
Tests the stated empty or repeated-input boundary.
Explains the result using the supplied records and a concrete test.

Data analyst
Junior
20 min

Measurements with zero and missing

Implement measurement_totals(rows). Each row has station and reading. Return a dict from station to {count, total}; count only non-None readings, preserve recorded zero, and include a station with only missing readings as count 0 and total None. Don't modify rows.

Start interview

Skills and rubricGrouping with explicit missing values

Initialize each station independently, then distinguish None from every measured number.
Tests the stated empty or repeated-input boundary.
Explains the result using the supplied records and a concrete test.

Test automation engineer
Junior
20 min

Debug notes that leak across requests

Repair append_note(note, notes=None). Return a new list containing prior notes followed by note. Repeated calls without notes must start fresh, and passing a list must leave that original list unchanged. The draft uses a mutable default and appends in place.

Start interview

Skills and rubricMutable defaults and caller state

Use None as the default and copy caller-owned collections before adding an item.
Tests the stated empty or repeated-input boundary.
Explains the result using the supplied records and a concrete test.

Data engineer
Mid
30 min

Quarantine malformed shipment lines

Implement parse_shipments(lines). Each line is code|units. Strip both fields; a valid line has exactly two fields, a nonempty code and a nonnegative integer units token accepted by int. Return {rows: [(code, units), ...], rejected: [one-based line numbers]}. Retain valid zero quantities. A malformed line must not stop the batch.

Start interview

Skills and rubricParsing with line-level diagnostics

Validate the record shape, handle expected conversion failures and retain a rejection audit.
Handles the revised requirement and preserves caller-owned state.
Reconciles the output against an independently chosen boundary example.

Data engineer
Mid
30 min

Reconcile a large reservation export

Implement enrich_reservations(reservations, owners). owners contains {id, name} rows with unique ids. Return a new list of {reservation, owner} in reservation order, using Unknown for absent or None owner_id. An owners row with id None is not a business key. Preserve every reservation. Explain how an indexed lookup avoids scanning owners for every row.

Start interview

Skills and rubricIndexed lookup and cardinality

Index valid owner keys once. Preserve source order and make unmatched-key handling explicit.
Handles the revised requirement and preserves caller-owned state.
Reconciles the output against an independently chosen boundary example.

Data analyst
Mid
30 min

A revised UTC reporting window

A service analyst wants event ids for a reporting window. Inspect the offset-aware timestamps and describe the window boundaries before implementing the filter.

Start interview

Skills and rubricOffset-aware half-open windows

Compare parsed offset-aware datetimes. Include the lower boundary and exclude the upper boundary.
Handles the revised requirement and preserves caller-owned state.
Reconciles the output against an independently chosen boundary example.

Software engineer
Mid
30 min

Review a replayed ingestion batch

Review apply_events(balance, events, seen). Each event has id and delta. Apply each unseen id once, even if repeated within this batch. Return (new_balance, sorted_seen_ids). Preserve the caller’s seen set and event rows. Replayed ids must not change the balance.

Start interview

Skills and rubricIdempotent event application

Copy the known ids and add each accepted event before considering the next event. The balance and id record need one persistent transaction in a deployed service.
Handles the revised requirement and preserves caller-owned state.
Reconciles the output against an independently chosen boundary example.

Data engineer
Senior
40 min

Bounded batches from a single-pass stream

Implement batch_records(records, size). Return a list of nonempty lists of at most size items, preserving order and the final partial batch. records may be a one-pass iterator. Raise ValueError for size <= 0. Explain what would change if the caller wanted lazy output rather than a collected list.

Start interview

Skills and rubricIterator batching and tail preservation

Accumulate one batch at a time and retain its final tail. The returned collected list still holds the full output; a lazy variant would yield batches.
Defends correctness, complexity and failure behavior with independent tests.
Defines a measurable runtime and memory check without unsupported speed claims.

Test automation engineer
Senior
40 min

Independent tests for a finished range helper

Implement audit_pages(candidate). It must return a list of case labels that candidate fails, in this order: empty, partial, exact, invalid. Test candidate(total, size), which should return the number of pages needed for total records; total 0 needs zero pages; size <= 0 raises ValueError. Use cases (0,4), (9,4), (8,4), (1,0). Treat an unexpected exception as failure, but a ValueError passes only invalid. Don't derive expected values using candidate itself.

Start interview

Skills and rubricBoundary tests and exception contracts

Use independently chosen examples and explicit exception expectations. A completed function still needs boundary evidence.
Defends correctness, complexity and failure behavior with independent tests.
Defines a measurable runtime and memory check without unsupported speed claims.

Software engineer
Senior
40 min

Deterministic pipeline dependency order

Implement schedule_tasks(dependencies), a dict mapping every task to its direct prerequisites. Return a valid task order, repeatedly choosing the lexicographically smallest currently ready task. Raise ValueError for a missing prerequisite or a cycle. Don't mutate the input. Disconnected tasks are allowed.

Start interview

Skills and rubricDependency validation and deterministic ordering

Track unmet dependencies, select only ready tasks and break ready-node ties explicitly. Exhausted readiness with tasks left means a cycle.
Defends correctness, complexity and failure behavior with independent tests.
Defines a measurable runtime and memory check without unsupported speed claims.

Choose your character

Python interviews

Timed assessment

Practice interviews

Stable participant identifiers

Measurements with zero and missing

Debug notes that leak across requests

Quarantine malformed shipment lines

Reconcile a large reservation export

A revised UTC reporting window

Review a replayed ingestion batch

Bounded batches from a single-pass stream

Independent tests for a finished range helper

Deterministic pipeline dependency order

Interview Hub access

Your Work Gym account