Chii chaitika
Databricks inoti yakawana mabhugi manomwe mumidziyo yemukati inoshandiswa neAI vamiririri vekodha uye mamwe mafambiro ebasa. Iyo kambani inofungidzira kutadza kwakakonzeresa madhora mazana mana nemakumi mapfumbamwe nemakumi mapfumbamwe emadhora mumatokeni akaraswa uye angangoita maawa gumi nemaviri enguva yekumirira mumiriri wepagore, iyo yainokoshesa inosvika madhora miriyoni imwe chete.
Databricks inotsanangura kuferefeta kwemukati mumutengo weAI maajenti anoshandiswa kukodha uye mamwe mafambiro ebasa. Kambani iyi inoti vamiririri vayo vanowana hwaro modhi uye maseva eMCP anopa maturusi ekushanda nezvinhu zvakaita sehurongwa matanda, matafura ekushandisa, matikiti ekutsigira, uye wikis. Sezvo mashandisirwo akawedzera, Databricks yaifungidzira kuti yakundikana maturusi mafoni aigadzira mari yakavanzika nekuti vamiririri vaiwanzoedza kana kushanda vakatadza pane kumira.
Iyo kambani inoti Unity Gateway yakagadzira otomatiki OpenTelemetry traces yeMCP chishandiso chekukumbira. Iwo marekodhi aisanganisira maturusi mazita, nharo, zvikanganiso, token count, , uye musangano identifiers, maererano Databricks. Iwo anoteedzera akachengetwa mutafura, achibvumira kambani kubatanidza kutadza kwega kwega neakazoedzazve, kushandisa tokeni, uye nguva yekumirira. Databricks inoti Genie One wobva warega mainjiniya abvunze iyo data mumutauro wechisikigo pane kunyora SQL mibvunzo nemawoko.
Muhwindo rimwe chete remaawa makumi maviri nemana, Databricks inoti yakaona zvikanganiso zana nemakumi mana nemapfumbamwe ezvishandiso pazuva paJira neGoogle Drive kana Docs maseva. Iyo kambani yakati inofungidzirwa $499,000 mumitengo yepagore yechiratidzo uye 12,023 maawa epagore yekumirira nguva kune manomwe anodzokororwa tsikidzi. Mabviro makuru pane akanyorwa kutadza kwekutsvaga kweJira kuchisanganisira ndima : sevha yaitarisira tambo yakapatsanurwa nemakoma, ukuwo mumiriri akapasa JSON list. Databricks inoti kukanganisa kwakaitika ka535 pazuva uye yakatora avhareji yegumi nemaviri matende kuti adzoke kubva.
Kumwe kukundikana kwakataurwa kunosanganisira kushaikwa kweJira, ongororo isingatsigirwe_yekukurumidza, sarudzo dzenzvimbo dzeGoogle Drayiri dzisina kunaka, Google Docs isipo, uye mabytes-versus-string mismatch. Databricks inoti vamiririri vekodha vakaisa zvigadziriso pamasevha ezvishandiso mushure meGenie One yaburitsa rondedzero yezvikanganiso uye zvakapinza zvakavakonzeresa. Iyo kambani inoratidzira iyo yakazara maitiro ekutsvaga, kuverengera, uye kugadzirisa nyaya sekutora ingangoita awa imwe.
Kwakabva mashoko: databricks.com ↗
Nei zvichikosha
Iyo account inoburitsa isingaonekwe sosi yeAI yekushandisa mutengo: vamiririri vanopora kubva pakatyoka maturusi mafoni nekuyedza zvakare, kufungidzira, kana kuyedza dzimwe nzira. Inopawo zano kuti maseva ezvishandiso anofanirwa kugashira zvinonzwisisika mumisiyano inogadzirwa-modhi pane kubata yese kusawirirana sekukanganisa kufona.
Chinhu chepakati chinoshanda ndechekuti basa reAI-agent rinogona kutaridzika rakabudirira uchiri kusashanda. Databricks inoti yakajairwa mutengo dashibhodhi inogona kuratidza chete kuwedzera zvine mwero mukushandisa tokeni uye kuita kuti itaridzike seyakajairwa kushandiswa kwekukura. Tsanangudzo dzinobatanidza zvikanganiso, kuedzazve, , uye zvikamu zvinogona kuratidza kuti mari yekuwedzera inoratidza basa rinoita basa kana kudzokororwa kudzoreredzwa kubva mukukanganisa kwezvivakwa.
Mienzaniso yacho inopikisawo fungidziro yakafanana pamusoro pekuvimbika kwechishandiso. Databricks inoti dzimwe nhare dzainzi dzisiridzo dzaive dudziro dzine musoro dzenzvimbo dzakasununguka dzakataurwa. Array ndeye yakasikwa JSON inomiririra rondedzero yeminda, semuenzaniso, kunyangwe imwe sevha yakanyorwa kuti igamuchire tambo yakapatsanurwa nemakoma. Mumamiriro ezvinhu akadaro, tsime rinopokana, kutadza chikamu idambudziko rekuenderana mune chishandiso pane kungokanganisa modhi.
Izvi zvine basa sezvo masangano anopa vamiririri mukana kune mamwe masisitimu ekushanda. Kufona kwakakundikana kunogona kudya ma modhi tokeni, kunonoka kufambiswa kwebasa, kuwedzera mashandisirwo ezvivakwa, uye kuita kuti hunhu huve hwakaoma kuongorora. Bvisa mhosho meseji inogona kuderedza nguva yekudzoreredza, asi Databricks inoti mushonga wakasimba ndewekugadzira maturusi anobata zvinogoneka kupinza misiyano, kupa zvine musoro kusarudzika, uye kuramba nharo dzisingatsigirwe nenzira inopa mumiririri nhungamiro inobatsira.
Iyo sosi zvakare chigadzirwa account kubva kuDatabricks, ayo maturusi ari chikamu chemhinduro inosimudzira. Huwandu hwayo fungidziro yakavakirwa pane emukati maronda uye fungidziro pamusoro pegore negore mutengo uye kubereka. Iyo account haipe yakazvimirira odhiyo, yakadzama-yekutendeuka nzira yemadhora miriyoni 1.2, kana humbowo hwekuti mari imwechete yaizochengetwa mumasangano ane mabasa akasiyana, mamodheru, maseva ezvishandiso, kana mutengo wevashandi.
Interactive Mechanism: Iyo Inonyatsoshanda
Ongorora ari pasi tekinoroji kuseri kwekusimudzira uku uchipindirana.
crm_get_transaction(id='4092').An agent must create a draft calendar event for Tuesday at 2 p.m. Which evidence would establish the requested result?
Zvekutarisa zvinotevera
Nhamba dzekuchengetedza iDatabricks fungidziro kubva kune yayo yeagent zvikepe, kwete yakazvimirira yakaongororwa mhedzisiro. Humwe humbowo hwaizodikanwa pakuderedzwa kwemashure ekutadza, kana zvigadziriso zvakaramba zvichishanda nekufamba kwenguva, uye kuti nzira yacho inoshanda sei kune mamwe maajenti masisitimu, maturusi, uye nharaunda dzekushanda.
Iyo inonyanya kukosha yekutevera ndeyekuti Databricks yakayera mhedzisiro chaiyo mushure mekugadzirisa. Kwakabva kunotsanangura marara anofungidzirwa kusati kwagadziriswa asi haatauri mushure mekugadzirisa kukanganisa mwero, kuderedzwa kwetokeni, shanduko yekudzoreredza majana, kana yakasimbiswa mari yekuchengetedza. Zviyero izvozvo zvaizobatsira kusiyanisa kuongororwa kunonzwisisika kubva pakusimbaradzika kwekuita basa.
Mapoka anoongorora maitiro anofanirwawo kuongorora kuvimbika kweiyo data yekutevera uye miganhu yekuongorora. Tsime rinoti Unity Gateway inorekodha nharo dzezvishandiso, zvikanganiso, tokens, , uye chikamu IDs, asi haikurukure sampling, kushaikwa kwemateki, zvidzoreso zvekuvanzika, kuchengetedza, kana mabatirwo anoitwa nharo. Iwo madeti ane basa kana kuteedzera kunosanganisira matikiti ekutsigira, matanda, zvinyorwa, kana imwe bhizinesi data.
Mamiriro ekuwanikwa kwezvigadzirwa zvepasi anogona kukanganisawo kutora. Databricks inoti Unity Gateway inowanikwa uye tafura yayo yakabatana iri mubeta, asi sosi yacho haipe mitengo, miganho yebasa, zvinodikanwa zvekutumira, kana mhedzisiro yekufananidza kune mamwe masisitimu ekuona. Izvo zvakare hazviratidze kuti Genie One anogona kupindura akavimbika mibvunzo mimwechete pane zvekupokana trace schemas.
Zvakawanda, nyaya yacho inomutsa mubvunzo kune vanogadzira ajenti: inguva yakadii yekuchinjika iyo chishandiso chinofanira kugashira chisati chamanikidzirwa chinobvumira chinogadzira njodzi nyowani kana yechokwadi? Kushandura otomatiki zvinopinza kana kufuratira nharo dzisingatarisirwe kunogona kuderedza marara, asi kunogona kuvanza kukanganisa chaiko. Remangwana tekinoroji zvinyorwa kana yakazvimiririra kuyedzwa kunofanirwa kuratidza kuti aya magadzirirwo ekuenderana anosungwa sei, akaiswa, uye akasimbiswa mune akakwira-stakes workflows.