Ukwakha Udokotela Wamaterorogist Wezimo Ezibonakalayo Esebenzisa Ama-Agents we-Amazon Bedrock

Ukuhlanganiswa kwamandla we-AI akhiqizayo kuyindlela yokuguqula izinguquko ezimbonini eziningi. Yize imininingwane yesimo sezulu itholakala ngeziteshi eziningi, amabhizinisi athembela kakhulu kwidatha ye-meteorological adinga izixazululo eziqinile futhi ezinokalisiwe ukuphatha ngempumelelo nokusebenzisa lezi zinqubo ezibucayi futhi anciphise izinqubo ezibucayi futhi zinciphise izinqubo ezibucayi futhi zinciphise izinqubo ezibucayi futhi zinciphise izinqubo ezibucayi futhi zinciphise izinqubo ezibalulekile. Lesi sixazululo sibonisa ukuthi ungawakha kanjani i-semeteorologist ye-virtual enamandla ebonakalayo engaphendula imibuzo eyinkimbinkimbi yesimo sezulu ngolimi lwemvelo. Sisebenzisa izinsizakalo ezahlukahlukene ze-AWS ukuhambisa ikhambi eliphelele ongalisebenzisa ukuxhumana ne-API enikezela ngemininingwane yesimo sezulu sangempela. Kulesixazululo, sisebenzisa ama-Amazon Bedrock Agents.

Ama-ejenti we-Amazon Bedrock asiza ekuhambiseni ukuhamba komsebenzi futhi ashintshe imisebenzi ephindaphindwayo. Ama-ejenti we-Amazon Bedrock angaxhuma ngokuphepha kwimithombo yedatha yenkampani yakho futhi akhulise isicelo somsebenzisi ngezimpendulo ezinembile. Ungasebenzisa ama-Amazon Bedrock Agents e-Architect anction action schema ehambisane nezidingo zakho, ekunikeze ukulawula noma nini lapho umenzeli eqala isenzo esibekiwe. Le ndlela eguquguqukayo ikuhlomisa ukuze uhlanganise umthungo futhi ukhiphe logic yebhizinisi ngaphakathi kwensizakalo yakho oyithandayo ye-backunder insizakalo, ukukhuthaza inhlanganisela yokubumbana kokusebenza nokuvumelana nezimo. Kukhona nokugcinwa kwememori kulokhu kuhlangana okuvumela umuzwa oqondene nomuntu owenzelwe wena.

Kulokhu okuthunyelwe, sethula indlela ehlelwe kabusha yokusebenzisa i-ejenti enamandla ai ngokuhlanganisa ama-Amazon Bedrock Agents kanye nemodeli yesisekelo (FM). Sikuqondisa ngenqubo yokumisa i-ejenti nokusebenzisa i-logic ethize edingekayo kudokotela wezimpendulo ze-meteorologist ukuhlinzeka ngezimpendulo ezihlobene nesimo sezulu. Ngokwengeziwe, sisebenzisa izinsizakalo ezahlukahlukene ze-AWS, kufaka phakathi ama-AWS akhulisekile ekusingatheni ukuphela kwangaphambili, imisebenzi ye-AWS Lambda yokuphatha i-Logic, i-Amazon Cognito yokufaka ubuqiniso bomsebenzisi, kanye nokuphathwa kokufinyelela (i-IAM) ngokulawula ukufinyelela kumenzeli.

Ukubuka konke

Umdwebo unikeza ukubuka konke futhi uqokomise izingxenye ezisemqoka. Ukwakhiwa kwezakhiwo usebenzisa i-Amazon Cognito yokufakazelwa ubuqiniso komsebenzisi futhi ngenze lula ukuba indawo yokusingathwa kwesicelo sethu sangaphambili. Ama-ejenti we-Amazon Bedrock adlulisela imininingwane evela kumbuzo womsebenzisi emaqenjini ezenzweni, abuye anikeze imisebenzi yangokwezifiso yeLambda. Iqembu ngalinye lesenzo kanye nomsebenzi weLambda uphatha umsebenzi othile:

1. I-Geo-Coordinates – Izinqubo Zokuxhumanisa Kwendawo (i-Geo-Coordinates) ukuthola imininingwane mayelana nendawo ethile
2. Isimo sezulu – Kubutha imininingwane yesimo sezulu yendawo enikeziwe
3. Isikhathi Sosuku – Ithola usuku lwamanje nesikhathi

Izimfuneko

Kufanele ube nalokhu okulandelayo endaweni yokugcwalisa isixazululo kulokhu okuthunyelwe:

Thumela Izinsizakusebenza Zesisombululo usebenzisa i-AWS Cloudformation

Lapho usebenzisa ithempulethi ye-AWS Cloudformation, izinsiza ezilandelayo zithunyelwe (Qaphela ukuthi izindleko zizotholwa izinsiza ze-AWS ezisetshenzisiwe):

• Izinsizakusebenza ze-Amazon Cognito:
• Izinsizakusebenza zeLambda:
  • Sebenza – -geo-coordinates-
  • Sebenza – -weather-
  • Sebenza – -date-time-
• Ama-Agents we-Amazon Bedrock: Udokotela we-Virtual-Meteorologist
  • Amaqembu Ezenzo (1) – obtain-latitude-longitude-from-place-name
  • Amaqembu Ezenzo (2) – obtain-weather-information-with-coordinates
  • Amaqembu Ezenzo (3) – get-current-date-time-from-timezone

Ngemuva kokuthumela ithempulethi ye-Cloudformation, Kopisha okulandelayo kusuka ku Okuphumayo Ithebhu ku-Cloudformation Console ezosetshenziswa ngesikhathi sokucushwa kwesicelo sakho ngemuva kokuthi ifakwe kuma-AWS Aliclive.

• AWSRegion
• BedrockAgentAliasId
• BedrockAgentId
• BedrockAgentName
• IdentityPoolId
• UserPoolClientId
• UserPoolId

Sebenzisa uhlelo lokusebenza lwe-AWS AM AMLY

Udinga ukuthumela ngesandla uhlelo lokukhulisa usebenzisa ikhodi engaphambili etholakala ku-GitHub. Qedela lezi zinyathelo ezilandelayo:

1. Landa ikhodi engaphambili ye-AWS-AmAmed-frontind.zip kusuka ku-GitHub.
2. Sebenzisa ifayela le-.Zip ukuze usebenzise kabusha uhlelo lokusebenza nge-ALMFENCE.
3. Buyela ekhasini lokukhulisa futhi usebenzise isizinda esikhiqizwa ngokuzenzakalelayo ukufinyelela uhlelo lokusebenza.

Sebenzisa i-Amazon Cognito ngokuqinisekiswa komsebenzisi

I-Amazon Cognito insiza kamazisi ongayisebenzisa ukuqinisekisa futhi igunyaze abasebenzisi. Sisebenzisa i-Amazon Cognito kwisixazululo sethu ukuqinisekisa umsebenzisi ngaphambi kokusebenzisa uhlelo lokusebenza. Siphinde sisebenzise i-IDITY POOL ukuhlinzeka ngeziqinisekiso ze-AWS zesikhashana zomsebenzisi ngenkathi zisebenzisana ne-Amazon Bedrock API.

Sebenzisa ama-Amazon Bedrock Agents ukushintshanisa imisebenzi yohlelo lokusebenza

Nge-Amazon Bedrock Agents, ungakha futhi ulungiselele ama-ejenti ezizimele kuhlelo lwakho lokusebenza. I-ejenti isiza abasebenzisi bakho bokugcina izenzo ezisuselwa kwimininingwane yenhlangano nokufakwa komsebenzisi. Ama-ejenti ama-Orchestrate Ukuxhumana phakathi kwe-FMS, imithombo yedatha, izinhlelo zokusebenza zesoftware, nezingxoxo zabasebenzisi.

Sebenzisa iqembu lesenzo ukuchaza izenzo ezenziwa ama-Amazon Bedrock Agents

Iqembu lesenzo lichaza isethi yezenzo ezihlobene ne-Amazon Bedrock Agent enze ukusiza ukusiza abasebenzisi. Lapho ulungiselela iqembu lesenzo, unezinketho zokusingathwa kwemininingwane enikezwe ngumsebenzisi, kufaka phakathi ukufakwa komsebenzisi eqenjini lesenzo se-Agent, kudluliswa idatha kumsebenzi we-lambda logic, noma ukulawulwa kokubuyisa ngqo ngempendulo evumayo. Ngohlelo lwethu lokusebenza, sakha amaqembu amathathu ezenzo ukuthi anikeze i-Amazon Bedrock Agent le nsiza ebalulekile: Ukubuyisa izixhumanisi zezindawo ezithile, ukuthola imininingwane yesikhathi samanje kanye nesikhathi, kanye nokulanda idatha yesimo sezulu ngezindawo ezinikezwe. Lawa maqembu ezenzo anika amandla i-ejenti ukuthi afinyelele futhi acubungule imininingwane ebalulekile, athuthukise amandla ayo okuphendula ngokunembile nangokuqondakala emikhondweni yomsebenzisi ehlobene nezinsizakalo ezenzelwe indawo kanye nezimo zezulu.

Sebenzisa i-Lambda ye-Amazon Bedrock Action Group

Njengengxenye yalesi sixazululo, imisebenzi emithathu ye-lambda ihanjiswa ukusekela amaqembu ezenzo achazwe nge-Amazon Bedrock Agent yethu:

1. Indawo ixhumanisa umsebenzi we-lambda – Lo msebenzi udalwe yi obtain-latitude-longitude-from-place-name iqembu lesenzo. Kuthatha igama lendawo njengokufaka futhi kubuyisa izixhumanisi ezihambisanayo nobude. Umsebenzi usebenzisa insizakalo ye-geocoding noma database ukwenza lokhu kubheka.
2. Usuku nesikhathi umsebenzi we-lambda – kufakwe yi- get-current-date-time-from-timezone Iqembu lesenzo, lo msebenzi uhlinzeka ngolwazi lwamanje nolwazi lwesikhathi.
3. Imininingwane Yesimo Sezulu Umsebenzi weLambda – Lo msebenzi ubizwe ngu obtain-weather-information-with-coordinates iqembu lesenzo. Yamukela ukuxhumanisa kwe-geo kusuka emsebenzini wokuqala we-lambda futhi ibuyisa izimo zezulu zamanje nokubikezela kwendawo ebekiwe. Lo msebenzi we-lambda usebenzise i-api yesimo sezulu ukuyolanda idatha ye-meteorological esesikhathini.

Ngayinye yale misebenzi ye-lambda ithola umcimbi wokufaka oqukethe ama-metadata afanele kanye nezinkundla ezinabantu abavela ekusebenzeni kwe-API ka-Asment ye-Amazon Bedrock Agent noma amapharamitha. Imisebenzi icubungula lokhu okokufaka, yenza imisebenzi yayo ethile, bese ubuyisela impendulo ngolwazi oludingekayo. Le mpendulo isetshenziswa yi-Amazon Bedrock Agent ukwakha impendulo yayo embuzweni womsebenzisi. Ngokusebenzisa le misebenzi ye-lambda, i-Amazon Bedrock Agent ithola amandla okuthola imithombo yedatha yangaphandle futhi enze amakhono ayinkimbinkimbi, athuthukise kakhulu amandla awo ekusingatheni izicelo zomsebenzisi ezihlobene nendawo, isikhathi, kanye nemininingwane yesimo sezulu.

Sebenzisa ama-AWS akhulisa ikhodi yangaphambili

I-Alipelly inikeza indawo yokuthuthuka yokwakha okuphephile, okubukeka kahle kanye ne-web application. Onjiniyela bangagxila kwikhodi yabo kunokuba bakhathazeke ngengqalasizinda engaphansi. I-Alebhusayithi kahle futhi ihlangana nabahlinzeki abaningi be-git. Ngalesi sixazululo, silayisha ngesandla ikhodi yethu yangaphambili esebenzisa indlela echazwe ngaphambili kulokhu okuthunyelwe.

Ukuhamba kwesicelo

Zulazulela ku-URL enikezwe ngemuva kokudala uhlelo lokusebenza nge-ALMFENTEL. Lapho ufinyelela i-URL yesicelo, uzokwaziswa ukuthi unikeze imininingwane ehlobene ne-Amazon Cognito ne-Amazon Bedrock Agents. Lolu lwazi luyadingeka ukuthi luqinisekise abasebenzisi ngokuphepha futhi luvumele ukuphela kwangaphambili ukuthi uhlanganyele ne-Amazon Bedrock Agent. Inika amandla uhlelo lokusebenza ukuphatha izikhathi zomsebenzisi futhi lwenze izingcingo ezigunyaziwe ze-API ezinsizakalweni ze-AWS egameni lomsebenzisi.

Ungafaka imininingwane ngamanani owaqoqe kusuka ekuphumeni kwesitaki se-Cloudformation. Uzocelwa ukuthi ufake izinkambu ezilandelayo, njengoba kukhonjisiwe ku-skrini elandelayo:

• I-ID Yomsebenzisi
• I-POOL Pool Clientiod
• I-ID kamazisi
• Indawo
• Igama Lomenzeli
• I-ID ye-Agent
• I-Agent Alias ​​ID
• Indawo

Udinga ukungena ngemvume ngegama lakho lomsebenzisi nephasiwedi. Iphasiwedi yesikhashana yakhiqizwa ngokuzenzakalelayo ngesikhathi sokuhanjiswa futhi yathunyelwa ekhelini le-imeyili olinikele lapho sethula ithempulethi ye-Cloudformation. Ekuqaleni kokungena ngemvume kokungena ngemvume, uzocelwa ukuthi usethe kabusha iphasiwedi yakho, njengoba kukhonjisiwe kuvidiyo elandelayo.

Manje usungaqala ukubuza imibuzo kuhlelo lokusebenza, ngokwesibonelo, “Singakwazi yini ukwenza izombuzo namuhla eDallas, TX?” Ngemizuzwana embalwa, uhlelo lokusebenza luzokunikeza imiphumela enemininingwane ekhulumayo uma ungenza izoso le-barbecue eDallas, TX. Ividiyo elandelayo ikhombisa le ngxoxo.

Isibonelo Sebenzisa amacala

Nansi imibuzo eyisampula embalwa ukukhombisa amakhono wesazi sakho se-meteorologist:

1. “Sinjani isimo sezulu eNew York City namuhla?”
2. “Ngabe kufanele ngihlele iphathi yokuzalwa yangaphandle eMiami ngempelasonto ezayo?”
3. “Kuzoba iqhwa eDenver ngosuku lukaKhisimusi?”
4. “Ngingakwazi ukubhukuda olwandle eChicago namuhla?

Le mibuzo ikhombisa ikhono le-ejenti yokuhlinzeka ngemininingwane yesimo sezulu samanje, inikezela ngezeluleko ezisuselwa ekubikezelweni kwesimo sezulu, futhi sibikezele izimo zezulu zesikhathi esizayo. Ungabuza ngisho nombuzo ohlobene nomsebenzi njengokubhukuda, futhi uzophendula ngokusekelwe ezimweni zezulu uma lowo msebenzi ulungile ukwenza.

Hlanza

Uma uthatha isinqumo sokuyeka ukusebenzisa udokotela we-meteorologist, ungalandela lezi zinyathelo ukuze uyisuse, izinsiza zayo ezihambisana nazo zisetshenziswa kusetshenziswa amathuba amasha, kanye nokuthunyelwa kwezeluleko:

1. Susa isitaki se-Cloudformation:
   1. Ku-AWS Cloud Cloudformation Console, khetha Izithinca kufasitelana lokuzulazula.
   2. Thola isitaki osidalile ngesikhathi senqubo yokuhambisa (owabelwe igama).
   3. Khetha isitaki bese ukhetha Cisha.
2. Susa uhlelo lokusebenza lwe-AMALLYNDY kanye nezinsizakusebenza zalo. Ngemiyalo, bheka Ukuhlanza Izinsizakusebenza.

Ukugcina

Lesi sixazululo sikhombisa amandla okuhlanganisa ama-Amazon Bedrock Agents namanye ama-AWS Services ukudala umsizi wesimo sezulu ohlakaniphile, wokuguqula. Ngokusebenzisa ubuchwepheshe be-AI nefu, amabhizinisi angasebenzisa imibuzo eyinkimbinkimbi futhi anikeze ukuqonda okubalulekile kubasebenzisi bawo.

Izinsizakusebenza ezingeziwe

Ukuze ufunde kabanzi nge-Amazon Bedrock, bheka izinsiza ezilandelayo:

Ukuze ufunde kabanzi ngemodeli ye-Anthropic's Claude 3.5 Sonnet Model, bheka izinsiza ezilandelayo:

Mayelana nababhali

USalman Ahmed ngumphathi we-akhawunti we-applity ephezulu yezobuchwepheshe ekusekelweni kwebhizinisi le-AWS. Uyakujabulela ukusiza amakhasimende embonini yokuhamba nowokungenisa izihambi ukuklama, ukusebenzisa, nokusekela ingqalasizinda yefu. Ngothando lwezinsizakalo zokuxhumana kanye neminyaka yesipiliyoni, usiza amakhasimende ukuthi asebenzise izinsizakalo ezahlukahlukene zokuxhumana ze-AWS. Ngaphandle komsebenzi, uSalman uyakujabulela ukuthwebula izithombe, ukuhamba, nokubuka amaqembu akhe ayizintandokazi ezemidlalo.

Sergio Barraza Ingabe ukuhola okuphezulu kwebhizinisi okuphezulu ku-AWS, ukusiza amakhasimende wamandla aklanywe futhi asebenzise kahle izixazululo zamafu. Ngesifiso sokuthuthukiswa kwesoftware, uqondisa amakhasimende amandla nge-AWS Service Adoption. Umsebenzi wangaphandle, iSergio iyisigingini sensimbi ehlukahlukene edlala isiginci, upiyano, kanye nezigubhu, futhi futhi yenza futhi iphiko chun kung fung fu.

URavi Kumar ngumphathi we-akhawunti we-applity ophezulu ku-AWS Enterprise ukweseka okusiza amakhasimende embonini yezokuvakasha kanye nomoya wokungenisa izihambi ukuqondisa imisebenzi yabo yamafu emathul. Ungumphumela we-It uchwepheshe ngeminyaka engaphezu kwengu-20 yesipiliyoni. Esikhathini sakhe samahhala, iRavi ijabulela imisebenzi yokudala efana nokupenda. Ubuye athande ukudlala ikhilikithi nokuhambela izindawo ezintsha.

I-Ankush Goyal Ingabe i-Enterprise Support eholwa ekusekelweni kwebhizinisi le-AWS esiza amakhasimende ukuthi anikeze amandla okusebenza kwawo amafu ngama-AWS. Ungumphumela we-It uchwepheshe ngeminyaka engaphezu kwengu-20 yesipiliyoni.