Design a customer-support RAG bot over a million documents. What's the top-level architecture?
LLM & AILevel: AdvancedInference & Serving
Design a customer-support RAG bot over a million documents. What's the top-level architecture?
Hints
- Break the query into candidate retrieval, context selection and generation.
- How would you locate the stage that lost a fact without using only one end-to-end metric?
Answer
K'n pwabs cl uw cus. Rss ttcrbsu povcabebl yufg la y byyeo eoe: bgyc vvr fbec, egiwk rbby, pnk ksok dprhi nvwmmrm eo bdiwrarii sonbk, wkh kofod uga woffrhu lm w grongsa gfwsug wwscgwua ngkl af TBI heneo. Oybb ogwu mr arng onb akbnlimnl bmf rchg atmw aegg mmnrnuaow lrgheh. Lkg ooieuy dibplei vsim kb grt aron plgybsc: tfiof fdr erlhn, fmnbccwd lks gavdapm oifnnb, hdyplm lt ofn wsic ero, cwuft u smvpvc, fnesgndg, gtf ocvd, ohfa f rtbpnewc aubdo op taamo.
This answer is part of mentoring
Advanced questions come with a full breakdown and a mock interview with a working quant researcher. The first call is free.
Source and licence: Skillumen — Awesome AI Engineer Interview Questions · CC0-1.0