Python developer interview questions for remote roles

Technical questions

What you know and how you'd do the work, often with a live task or test.

1. What is wrong with def add_item(item, items=[]):?

   Why they ask

   The classic Python gotcha. It shows whether you understand mutability and when defaults are evaluated.

   How to answer

   Default values are evaluated once, when the function is defined, so every call that does not pass items shares the same list and it keeps growing. Use items=None and create a new list inside the function when it is None.

2. What is the difference between a list and a tuple?

   Why they ask

   A basic check that leads to hashing and immutability.

   How to answer

   Lists are mutable; tuples are immutable. Tuples of hashable items can be dictionary keys or set members; lists cannot. Use tuples for fixed records and lists for collections that change. Mention named tuples or dataclasses for readable records.

3. What is a generator, and when would you use one?

   Why they ask

   Generators matter for memory use with large data.

   How to answer

   A generator produces values lazily using yield, keeping its state between calls, so it does not build the whole sequence in memory. Use it to stream a large file line by line or process database rows in batches. Generator expressions use parentheses instead of square brackets.

4. Write a decorator that times how long a function takes.

   Why they ask

   Decorators appear throughout Django, FastAPI and testing code.

   How to answer

   A decorator takes a function and returns a wrapper. Use functools.wraps to keep the original name and docstring, and time.perf_counter for timing.

   Show a sample answerHide the sample answer

   ```python
   import functools
   import time

   def timed(func):
   @functools.wraps(func)
   def wrapper(args, *kwargs):
   start = time.perf_counter()
   try:
   return func(args, *kwargs)
   finally:
   elapsed = time.perf_counter() - start
   print(f"{func.name} took {elapsed:.3f}s")
   return wrapper
   ```
   I used something like this to find which step of a nightly report job was slow before moving it to Celery.

5. What is the GIL, and how do you handle CPU-bound work?

   Why they ask

   Shows you understand Python's performance limits.

   How to answer

   In the standard CPython build, the Global Interpreter Lock lets only one thread run Python bytecode at a time, so threads do not speed up CPU-bound Python code, though they help with I/O-bound work. For CPU-bound work use multiprocessing or a process pool, push the heavy work into libraries written in C such as NumPy, or offload to a worker. CPython also has an optional free-threaded build without the GIL, but most production code still runs on the standard build.

6. When would you use asyncio instead of threads?

   Why they ask

   FastAPI and many modern services are async.

   How to answer

   asyncio suits many concurrent I/O operations, such as calling several APIs or handling many slow connections, with less overhead than threads. Everything in the call path must be async-aware; one blocking call (like requests or time.sleep) stalls the event loop. Use httpx or aiohttp for HTTP and run blocking code in a thread executor.

7. What is the N+1 problem in Django and how do you fix it?

   Why they ask

   The most common Django performance bug.

   How to answer

   Looping over a queryset and accessing a related object triggers one extra query per row. Use select_related for foreign keys and one-to-one relations (a SQL join) and prefetch_related for many-to-many and reverse relations (a second query). Find it with Django Debug Toolbar or by asserting query counts in tests with assertNumQueries.

   Show a sample answerHide the sample answer

   An orders admin page at my last job took 6 seconds. Debug Toolbar showed over 300 queries because each row loaded its customer and its items. Adding select_related('customer') and prefetch_related('items') brought it to 3 queries and under half a second.

8. How does dependency injection work in FastAPI?

   Why they ask

   A core FastAPI feature used for auth, database sessions and settings.

   How to answer

   You declare a parameter with Depends(some_function), and FastAPI calls that function per request and passes the result in. Dependencies can depend on others, can use yield for setup and teardown (such as opening and closing a database session), and can be overridden in tests with app.dependency_overrides.

9. How do you write tests that call an external API?

   Why they ask

   Tests that hit real services are slow and flaky.

   How to answer

   Isolate the API call behind a function or client class, then mock it in unit tests with unittest.mock or pytest-mock, or use respx or responses to fake HTTP. Keep a small number of integration tests against a sandbox. Use pytest fixtures to set up shared data.

10. What is the difference between is and ==?

    Why they ask

    A quick check for understanding identity versus equality.

    How to answer

    == compares values by calling eq; is checks whether two names point to the same object. Use is for None, True and False checks, and == for values. Small integer and string caching can make is look like it works for values, which is why it is a trap.

Behavioural questions

Past situations. Answer with a real example: what happened, what you did, what changed.

1. How do you keep code quality high on a team?

   Why they ask

   Python's flexibility can produce messy codebases.

   How to answer

   Ruff for linting and formatting, type hints with mypy or Pyright in CI, pytest with sensible coverage, small pull requests and clear review comments. Give an example of introducing one of these.

2. Describe a time you automated something that saved your team time.

   Why they ask

   Automation is one of Python's biggest selling points.

   How to answer

   Explain the manual process, what you built, and the time or errors saved.

   Show a sample answerHide the sample answer

   Our finance team in Lagos spent two days each month matching bank statement exports to invoices in Excel. I wrote a Python script with pandas that matched by amount, date and reference, and flagged the leftovers for review. It cut the work to about two hours, and the flagged list caught a duplicate payment in the first month.

Situational questions

Hypotheticals about problems you'd meet in the job.

1. How would you process a 5GB CSV file on a laptop with 8GB of RAM?

   Why they ask

   A practical data question that tests memory awareness.

   How to answer

   Do not load it all at once. Stream it with the csv module, or read it in chunks with pandas (chunksize), or use Polars' lazy scanning or DuckDB, which can query the file directly. Aggregate as you go and write results incrementally.

2. A Celery task sometimes runs twice and sends duplicate emails. What do you do?

   Why they ask

   Distributed task queues usually deliver at least once.

   How to answer

   Make the task idempotent: record that the email was sent (with a unique key) in the database and check before sending. Review retry and acknowledgement settings, such as acks_late and visibility timeouts, which can cause re-delivery when tasks run longer than expected.

Remote work questions

How you'll work from Nigeria with a team elsewhere: hours, power, internet, communication.

1. How do you handle tasks when nobody on your team is online to answer questions?

   Why they ask

   Async work across time zones is normal.

   How to answer

   Write a clear question with context and options, make a reasonable assumption and note it in the pull request, and keep a second task ready.

2. Have you used AI coding assistants, and how do you check their output?

   Why they ask

   Teams want developers who use AI tools carefully, especially in Python, where wrong code often still runs.

   How to answer

   Say how you use them (boilerplate, tests, unfamiliar libraries) and how you verify: run tests, read every line, check library versions and never paste secrets or private code into tools the company has not approved.

Remote interview tips

Questions people ask