Pick the operational rules that prevent stale-embedding bugs in production

Attempt it

Quick reference

TL;DR

The four sound rules track model identity, hold the same model on both sides, re-embed on upgrade, and pin the snapshot. The trap option is 'let new docs use a newer model'.

Memory aid

Sign in for the mnemonic that makes this stick. See Pro pricing.

Easy to grasp

Think of it like a soccer team that switches uniforms mid-season. Half the players wear the old jersey, half wear the new one. From the stands you can no longer tell who is on which team because the colors no longer line up. To keep the game readable you need everyone in the same jersey at any moment, you need to write down which season's jersey it is, and when you change uniforms you swap all of them at once, not one by one. A search index works the same way. Same tool on every side, label what you are storing, swap everything together when you upgrade.

Key concepts

Embeddings Vector databases

Detailed answer & concept explanation

~3 min read

Everything you need to truly understand this topic: intuition, mechanics, step by step explanation, code, formulas, and worked example. Click to expand.

Everything you need to truly understand this topic: intuition, mechanics, step by step explanation, code, formulas, and worked example.

Sign in for 5 free unlocks/day or go Pro for unlimited. See Pro pricing.

Quick scan · 2–4 min

Everything important, quickly.

3 min: geometric invariant + model_id metadata + query-index alignment + parallel-index migration + snapshot pinning + drift detection.

Real-world examples

Real products, models, and research that use this idea.

Pinecone, Weaviate, and Qdrant all support per-vector metadata; production RAG teams in 2026 store model_id and model_version on every upsert.
OpenAI exposes dated model aliases like text-embedding-3-small-2024-01-01 so teams can pin a stable snapshot in production.
Cohere's embed-v4 migration guidance in 2026 recommends the parallel-index pattern with dual-write and shadow-test stages.

Interviewer follow-ups

What an interviewer would ask next. Try answering before peeking at the approach.

QWhat does the model_id metadata buy you operationally?

Sign in or go Pro to see the approach for this follow-up.

QHow would you detect that a vendor silently bumped their model snapshot?

Sign in or go Pro to see the approach for this follow-up.

QWhen does pinning the snapshot stop being worth it?

Sign in or go Pro to see the approach for this follow-up.

Don't say this

Red flags and common mistakes that signal junior thinking. Click to expand.

Most common mistake

Letting new documents use a newer embedding model while older indexed documents stay on the old one, on the theory that 'newer is always better'.

Sign in or go Pro to see all red flags and common mistakes. See Pro pricing.

60-second cheat sheet

The night-before-the-interview bullets. Scan these on the way to the call.

Sign in for 5 unlocks/day or go Pro for unlimited revision sheets. See Pro pricing.

References

Primary sources. Skim if you want the original framing.