Izinkampani zabo bonke osayizi zisheshisa ukwamukelwa kwama-ejenti obuhlakani bokwenziwa, zikhangwa yisithembiso sokusebenza kahle okukhulu kanye nokunciphisa izindleko zokusebenza. Kodwa lo mjaho ufihla ingozi okukhulunywa ngayo kancane: ingxenye enkulu yale misebenzi akulindelekile ukuthi iqhubeke iphila iminyaka emibili ezayo, idle utshalomali ngaphandle kokuletha imbuyiselo ethenjisiwe.
Lesi sixwayiso sisekelwa idatha evela ezikhungweni ezimbili ezihlonishwa kakhulu emkhakheni wezobuchwepheshe. UGartner ubikezela ukuthi ama-40% ezicelo zebhizinisi azoba nama-ejenti obuhlakani bokwenziwa agxile emisebenzini ethile ekupheleni kuka-2026, uma kuqhathaniswa nangaphansi kuka-5% ngo-2025, futhi ngesikhathi esifanayo ubikezela ukuthi amaphrojekthi e-AI angaphezu kuka-40% azokhanselwa ekupheleni kuka-2027 ngenxa yokwenyuka kwezindleko, inani lebhizinisi elingacacile, noma ukulawulwa kwezingozi okunganele. Ucwaningo lweDeloitte luqinisa lesi simo ngolunye uhlangothi: cishe izingxenye ezintathu kwezine zezinkampani zihlela ukwamukela i-AI ye-ejenti eminyakeni emibili ezayo, kodwa ama-21% kuphela azo athi anemodeli yokuphatha evuthiwe yala ma-ejenti.
Ngokusho kukaCarlos Pisani, uchwepheshe wezakhiwo zezinhlelo kanye nomlingani osungula i-ArcH, izinombolo zichaza iphethini ayibonayo ekusebenzeni, isikhathi eside ngaphambi kokuba noma yimuphi umbiko uyiqinisekise. “Inkinga cishe ayikaze ikhethwe imodeli ye-AI. Ukubeka i-ejenti ezimele phezu kohlelo olungakaze lube nemininingwane ecacile yobuchwepheshe noma ukwakheka okuqinisekisiwe. Isendlalelo ngasinye esisha esengezwe kulesi sisekelo sandisa isikweletu sobuchwepheshe esikhundleni sokuxazulula noma yini,” kusho yena.
Ngokusho kukaPisani, kukhona ushintsho ekuziphatheni kwezinkampani ezingakafiki endleleni ezakha ngayo isofthiwe. “Imakethe ivuthiwe ngendlela ecabanga ngayo ngebhizinisi. Isidingo sokusebenza kahle kanye nokuzinza kwezezimali sikhulile; lokhu akusaphikiswana ngakho. Inkinga ukuthi lokhu kuvuthwa akukafinyeleli ekuthuthukisweni kwezinhlelo. Izinkampani eziningi zithuthukise imodeli yazo yebhizinisi futhi ziyaqhubeka nokuthuthukisa isofthiwe ngendlela endala, ngaphandle kwemininingwane kanye nokuqinisekiswa kwesakhiwo. Yilapho isikweletu sobuchwepheshe sizalwa khona, esibe sesivimbela ngqo ukusebenza kahle okudingwa yibhizinisi ngokwalo.”
Ngokusho kokuhlaziywa kukaPisani, idatha yeDeloitte iqinisekisa ukuxilongwa ngeso lokuphatha. "Iqiniso lokuthi izinkampani ezingama-21% kuphela ezinemodeli yokuphatha ama-ejenti avuthiwe akuyona inkinga yokuthobela imithetho, kuyisibonakaliso sokwakheka okubuthakathaka. Ukuphatha ama-ejenti kusikisela ukuthi kukhona okuthile okuhlelwe ukubusa. Lapho uhlelo luzalwa ngaphandle kwemininingwane, alukho uhlaka lokuphatha olungaluxazulula kamuva."
Ukuze kuncishiswe ingozi yokuba phakathi kwamaphrojekthi angu-40% akhanseliwe, uPisani uthi imininingwane yobuchwepheshe kufanele iphathwe njengezinqumo zebhizinisi, hhayi njengezinyathelo zobuchwepheshe okufanele zeqiwe. “Ngaphambi kokubuza ukuthi iyiphi i-ejenti ye-AI okufanele isetshenziswe, umbuzo ofanele uwukuthi ngabe ukwakheka okuyisisekelo kungasingatha leso sinqumo. Lokhu yilokho esikubiza ngokuthi i-Spec-Driven Architecture (SDA): akukho sinqumo sobuchwepheshe esenziwa ngaphambi kokuba kube nokucaca mayelana nokuthi yini okudingeka ixazululwe nokuthi izoqhubeka kanjani ngokuhamba kwesikhathi. Izinkampani ezeqa lesi sinyathelo azongi isikhathi; zihlehlisa ukukhanselwa kwephrojekthi yazo.”
