Ai Coding And Product Systems hard
Your AI reviewer splits PRs into one chunk per file. It missed a bug in a 4,000-line, 60-file PR where a function signature changed in one file and broke callers in three others. How should chunking change?
- A Chunk by changed symbol, attaching each changed function's callers and callees from the code graph, even across file boundaries.
- B Send the full 4,000-line diff in one prompt so the model sees every change together and can reason across all files at once.
- C Chunk by fixed 300-line windows with 50-line overlap, which ensures adjacent context spans file boundaries in the diff output.
- D Review each hunk independently against retrieved team rules and concatenate the findings, since most real bugs are local to one hunk.
Answer: A. Chunk by changed symbol, attaching each changed function's callers and callees from the code graph, even across file boundaries.
**Symbol-centric chunking** follows real dependencies, at the cost of graph-building work and overlap between chunks. One giant prompt (B) is tempting, but attention spreads thin across 4,000 lines and the false positive rate goes up. It only works for small diffs. Fixed windows (C) cross file boundaries arbitrarily, not along call edges.
One system design question a day, with spaced repetition so the ones you miss come back.
Get the appRelated questions
- Your assistant uses a classifier to route queries to a small or large model, which saves 60% of GPU spend. Analysis shows about 20% of hard queries are misrouted to the small model, which then gives confident wrong answers, and regenerate clicks on those threads have doubled. What should you do?
- Your workspace AI retrieves the top-k chunks from a shared vector index and passes them to the model with ACL metadata and a system prompt saying 'only use documents this user can access.' A user just received an answer quoting a private HR page they cannot open. What is the correct fix?
- Your support bot answers billing questions. One user has asked the same refund question three times, and retrieval confidence for that topic is low. What should the bot do next?
- Your AI review system receives a 4,000-line pull request spanning 60 files. Sending the whole diff in one prompt yields shallow, generic comments, and real bugs in files near the middle of the diff are missed. How should you structure the review?
- In your ChatGPT-style assistant, power users keep single threads running for weeks, reaching about 150K tokens. Cost per turn keeps climbing, latency is rising, and the model increasingly misses facts from early in the thread. What should you do?
- Your workspace AI filters retrieval using ACLs stored in the vector index, which is rebuilt nightly. An admin removes an employee from a confidential project page, but for the rest of the day the employee can still get AI answers drawn from it. What is the right fix?