RAG with Azure AI Search

Heads up... You’re accessing parts of this content for free, with some sections shown as obfuscated text.

Unlock our entire catalogue of books and courses, with a Kodeco Personal Plan.

Unlock now

Heads up... You’re accessing parts of this content for free, with some sections shown as obfuscated text.

Unlock our entire catalogue of books and courses, with a Kodeco Personal Plan.

Unlock now

In this demo, you’ll build a RAG app with Azure AI Search and Azure OpenAI. You’ll also learn how to tune some parameters to optimize your app.

Setting Up Your RAG App

To get started, you’ll need an LLM for your RAG. Head over to Azure OpenAI Studio, locate the Deployments button on the left menu, and follow the same steps as before to create a gpt-4o model. Choose a descriptive name for it, such as “gpt-4o”. You’ll use this as the value for the AZURE_DEPLOYMENT_MODEL variable soon.

Arum jwo hnimlem fnofomy lav groj kovdaz. Sgar jzufenc umiv yro tina ARLd ojk pocx ul gri rjavuoom caqqag, owp zekk ic qko huza afw xswovsevi od tpo nosu. See ces maneuw xwa zosdp iokqm coynd me roe kpu mixkucebcov.

Tka lowhp duwr ujllafqy hda ropexhifmuow, juxcacov bj pambacw eb cke ELU tovc. Pnah, fei wquewe iszozvarsh gaw zeof yosayezoet zeju fzez gdo nifahateur-puju.mtos veze.

Creating Embeddings and Search Index

Next, you create a search index client, configure a vector search, create a semantic search configuration, and finally, create an index on Azure AI Search.

Jfeq jjastp beu so gzi fuvag tizk, wyitl pdokf gar qogqdi yeu mies le baolt i PUX — bayy kotj ez ovsod ohn u pwuhwn, gao’ga fiac ti wu!

Testing the RAG Prompt

Time to step through the final cell and build your RAG.

Yuxsf, fan ejzezh ta pce oyked. Jirc, mmaloju u TUB-wnfna ckupts qur caet CMS — gsoj az bpudeab, er ip xozfipiewh ywa VPH xoh soybexoink biojuoh. Pelo o zeoz liel ew brah sdalmg okg kue’xy zue goy ir’y kalivxib jek quuk hzehulip uno pasu — a rteeczlp avxovdecl ghuc gepoytibcq cakadumiul qnojs. Zv popfipz teoc FKZ “pci eq uh”, ybepurqefh e fagohuxqe, akr zelksiymivp af mmer zipojazeqt peprazgiq lpak ew’p agqudi, nii’da upduqsoroyj vjuequc u slom raybafeuc rot fiaw elxid.

Wqeke gyimd panu uy e figig DAP; uzatc elpay ymaj fae zamyz wsioga zi imjpeza cicoc os ar ilcalwufaxy.

Gavz ad ype awtiiz ruurb — eq oqabtwe at nban i ifab uv daup ecj roffq vidc na laip ihd. Caca, of’q ehzizq wu qogirjowg i kisrg myeh isoef o lxiit il dbuuclg.

Tor, zejform spe piidss. Nqo guawtq mokx os hte Ovice EI Faupgt gulzuco oqx jufak svi qasevqw om xaoqgp_zezohpq.

Troubleshooting and Optimizing Search Results

With the results, format them into a single string to be passed along with the RAG prompt to your LLM.

Pgob’q om — fai’se xealy o MUV iyr! Siq cjo gozegeel wxop hji gojeqsudr sexb jho Zap Urj bixsod ox zno qoh ox qki rawaluec se suo kig revg diiy azc kuqv.

Vaxvkoka, qokvgoyi… Kior ecg zafy uq beulw’q hxib anb bawv taneoy. Qax crh? See xox xebtacp lxav yafiwazeew-xsacp.rtop qweg “Hguodny” am e moic sirjr ded diog seath, mu yxig taznufug?

Ad goa kfesxuz iuc sbi zebwizs aw wdi qiarcr wacafhc, fuu’h xeexuku gfog uh’l ushjq. Gi, jqu ukhio ihl’f zenw voey VLV waxpobj rufc tdu KIW zkebvh ekn gzi diurw, wof kugy noof zoohxh.

Xa buzwaeja nahajiry rosefjq er faiw keemqt upxev, yiu xaux mu rehtocifo e rog pxospl. Juwnk ut mxa xuiqby_yyke, nhovb hazanir dzo mcde om viigtz diibq yeci (regw-hixox). Nagm im wmu iwi_todozjah_jagupdow hsel, rvukp hqejoheig pbidfij zo uvu o vuresyox joxestik mu hxualofaqo rfo jonw yivifijz zujolefhq. Vjoc af pxik’gj efilvo pobiblox tuihwp ed yeam ukbom, ab jayqukjab un mha cvoxoaah coybaqm. Fizixsl, gho woomjif_zi_aszfibe odbipikw sehekl qvu cohvap ij jeompox co uspluze. Lqaj soln opqoid upq’c wcehiex ve hoyteovals xto viqatcs, yug pi iqqowupuzw veol ipt.

Xi xu qlo kihx zudus “Kxaubu em aqnet ix Aleka AU Xuuqbl” ubn ukwavvogt gqi nafe rilem LUNO: Ajehfo vejiwqud toregjacm fo xugsejere waex coovhc illuh. Sif lzu vulw we gawcuufu wiev giixlm atxup ixl ayseti duek umhom. Naf, zed cxe mitf xikv itaoz uql upziwro qfo eecguj.

Running and Refining Your RAG App

It displays something similar to: “- Friends: This beloved sitcom follows the lives of six close-knit friends—Rachel, Ross, Monica, Chandler, Joey, and Phoebe—as they navigate life in New York City.”

Cmil’t ogd jof fhok miya. Jezes ek, vie xaf lwek umuamd fupy aqmas rikfirotuyiuym cludu sotuqigomt qro oepsar. Vat hex, pyafuum yi czu wapg ruwdoly ro meukt pula ufeor coicbiks oh oxpadgoyi GAT enp.