Ai Coding And Product Systems medium
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?
- A Once a thread passes a token budget, summarize older turns into a running summary and keep the most recent turns verbatim.
- B Send the full thread on every turn, since a model with a large enough context window attends to all past messages equally well.
- C Truncate to the last N messages only, dropping older turns entirely so that the prompt size stays constant and predictable.
- D Store the conversation without limits and rely on prompt caching to make resending the full history on every turn effectively free.
Answer: A. Once a thread passes a token budget, summarize older turns into a running summary and keep the most recent turns verbatim.
**Compaction** trades some fidelity on old details for bounded cost and latency, and threads past roughly 50K-100K tokens need it. Prompt caching (D) is tempting, but it only discounts prefill. It does not remove the context limit or fix attention dilution over a huge prompt. Plain truncation (C) is acceptable for short, casual chats where early context rarely matters.
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?
- 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?
- Free users consume 70% of your assistant's GPU hours but bring in about 5% of revenue through conversions. Leadership wants to keep the free tier useful enough to drive upgrades. Which policy fits?