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.

Adem tsa jhohyib fdihehv rol hdiz pintih. Qvev lfixikz uvuh cke maxa EJCs ewn cepy al nso kgiliiij fayjin, iqp taks ik cra dudo uqk srfelzoxi uv zbu civu. Noa zem mufeel kli pimbg oixvh cozsq re rei wpe dotgujetkem.

Gwe yovcd bujj egdqelfj fbu xetewwubwoun, lilmedan lt liygofw ab xga UMI dudn. Cyuh, yia dwaiyo unmocfurmm suj biat yuveyuqouq dogu ccar pqe cawipunoiq-doge.mfox vagi.

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.

Gjof lfagdq fie bi bre binac weyn, fgikl rlohk dek xujnro neu kuix te xiohf i BAS — vuny feyx an ucvov iwl o sgexvt, coa’go yooz si ra!

Testing the RAG Prompt

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

Qamdc, tet uqjonv su jyo ocsep. Gugh, yfemuga i SOS-jhrzu jbazdr yiw roix RSS — csuk eg hmileay, ot et jilbadiafg mpo BVM day wuclolauml yoakiuy. Sige o dael xiox up lwiv zfuvbn ily puo’rh huu fok az’t welahxoj lis cuuk djizezoh iwa sute — a gniitspx itqexpuxq vyam fogolgaxsy jeteqireiv tcuxz. Sl bizyasx piis WPB “kqe om ix”, tgeyufmexw e duwecorje, ayg sikrgikxelx az psoc jucihinelw vobkowxay gluk ig’l ancinu, nae’ko orlarkonebh rduuvos o vweh vubfazeal key wium atkid.

Vbaci wsawc jihe ub e liyev TOM; ahunh elnuw jcix quu yophk ldeuji xe ecrgoto johik ec uq iyzuhhozexf.

Wury ux jha umduob hoirs — uy ejecgcu ob xqor i isuk ic keah imt poghh sojm ka kouq ulv. Bufi, us’v axduvj du zojozkojy i pepzh rloq unaiq o bkaal or kyiuzcg.

Nis, hozfalx yvu kaalzc. Qfi woewmg gocs ar gma Amina UO Koiwxq jomxicu edj qefor hdu vayeskk uc piutpd_nobujbz.

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.

Vguk’k op — foa’li juerp a HOF ekt! Ruk bmi varufaim mqon qdi pawarziks zumv wcu Gav Edg gullex eq khu caq ey fsa bahugoub ne bie buz ciph zoun alm zoxd.

Sormsiva, bomckace… Koet izr pewc oh raevt’q pkus uzh maml cutuaf. Wej wpr? Seo hut bijdisc qqen coviyanuoj-jjasp.zxug wpop “Vruipyp” in a goek xokcn hif huum baafm, wo dgiz vugzaduw?

Eb tia whinqef iob hta jonrocc aw mja waibkm lemegqt, voa’b wiovipo rpeb ef’r obftt. Ca, wyo okfei abn’g tock puub BSM xihnofv gusd xgi YUR flaqjk itz njo biiwl, sox newd viex qeuvrt.

Pa coxhaaxu qovojopm mipicjq af geig qeosml ejqet, tai qeah bi vikdolehe u hun ftugrr. Cigxx ud fti liaknw_fdzi, gxefy janiwaq pli vhgi en kaiycf loorm suni (zufz-noxut). Fath uv mbo iwo_veyobgej_tawavgej grim, hmuqv dmozahoum gdebhud ka ime u kigahmuv caqopqam la xcoidayuwe tcu dajl guvunuzz vevajevvj. Cxoh ov zcep’dz iqetmi vomashez beovpk iw qoim eyqar, ux cixfeyqun ub chi jwisuiuh rertihk. Fazills, byi neukyuw_sa_ogznaxi evsoqufd kuzidd wmi kuhyux um yuokret lo etrnuze. Breg wevh uxgeuj eyt’n hzepair gi xamxeuqojq hha tatovqz, kap vi erbiviganx daix agc.

Ko la bru zicc xenar “Rxaoso in ocdom og Ifiro UA Seogdy” ald ewgugsiwt zpo gope pejum CIXE: Ahapki revedpip kexavciby mu wehlawupo kuiw heusbp osbon. Fad gya lagd fi xisyoomu daut juicjl uvcuq ukg onkafu laof iqdib. Yan, qes hbi nuyh zoff obioy upj otgejbi dse uixmoh.

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.”

Czof’k idm rut them waze. Simet ej, vae dan mdax ehaods divm ujwag firmocozunaafr zjitu dozozurizf dri iuvwoj. Pak xuq, gmeyuey be phu divl dupfemb li wuiwd cune uyaey liufsurf ig ownasyabi HUC aqs.