Ngiyikhipha Njani Imitsetfo Yebhizinisi Kusistimu Yami Lendala nge-AI?

Ukhipha imitsetfo yebhizinisi kusistimu yakho lendala ngetinyatselo letingu-3: i-AI ifundza ikhodi futsi ibhale umtsetfo ngamunye njengemusho locacile lexhunywe kulayini lowuvela kuwo; umnikati welichubo uyawubuketa bese uyawucinisekisa noma uyawulungisa; bese wena uwucatsanisa naloko sistimu lekwenta ngempela kuma-invoyisi langempela. Lithulusi lelinjenge-AWS Transform yema-mainframe libhala imitsetfo njengetidzingo letingabuketwa, futsi kusimemetelo se-Thought Machine ne-Amazon sangamhla 29 Inyoni 2026 kukhipha kwendlula etigabeni letingu-4 letigcina ngekuvunywa ngumuntfu. Ungabhali likhodi lensha ngaphambi kwekutsi imitsetfo icinisekiswe.

Encenyeni 3 uphume neluhlu nesincumo sesikhashana kuyo yonkhe sistimu. Emasistimu loncume kuwakha kabusha noma kuwaguculela ngiwo ladzinga lesinyatselo kucala.

Kungani imitsetfo yebhizinisi iyintfo leligugu kakhulu kusistimu yakho lendala

Umtsetfo webhizinisi yindlela inkampani yakho lebala ngayo noma lencuma ngayo lokutsite: sesaphulelo sesikhasimende lesitsenga ngebuningi sibalwa njani, inhlawulo yekwephuta yengetwa nini, ngubani lovuma i-oda lekutsenga lengetulu kwemkhawulo lotsite, imali yesitolimende igcwaliswa njani. Lemitsetfo yakheke iminyaka leminyenti yetinguco, futsi leminyenti yayo ayizange ibhalwe kunoma nguyiphi incwadzi.

• Ekhodini: imibandzela nekubala lokwabhalwa ngumhleli eminyakeni leyendlulile ngesicelo semphatsi losashiye inkampani.
• Kusizinda selwati: emathebula emanani, emazinga nemikhawulo lokuntjintjwa ngumuntfu ngesandla angati kutsi yini lencike kuko.
• Etingcondvweni tebasebenti: imikhawulo lehlukile lesingatwa ngumcophi wetimali ngesandla ngobe sistimu ayikwenti loko.
• Kumaspredishithi: emafomula lavala tikhala tesistimu, njengobe sichazile ku-ngiliphi lishidi lelisistimu.

Nangabe wakha kabusha sistimu ngaphandle kwalemitsetfo, utawakha sistimu lensha lebala ngendlela lehlukile, futsi utawukubona loko kuphela nangabe likhasimende liphikisana ne-invoyisi. Imitsetfo ngiyo lekufanele ihambe; lulwimi lwekuhlela loludzala nje sitsebe kuphela.

I-AI iyifundza njani ikhodi futsi ibhale imitsetfo

Kukhipha imitsetfo bekungumsebenti wesandla lomudze: umhleli bekafundza ikhodi layini ngelayini futsi abhale loko lakuvisisile. Emathulusi alamuhla afundza yonkhe ikhodi, alandzelele kutsi inombolo ngayinye ivelaphi futsi iyaphi, bese abhala loko lakutfolile ngelulwimi lolucacile. Sibonelo sinye: i-Amazon itsi elikhasini le-AWS Transform yema-mainframe lensita ikhipha wonkhe umtsetfo webhizinisi kukhodi yemtfombo futsi iwulandzelele emsuka wawo, bese yakha tidzingo letilungele kutfutfukiswa leticembu lakho lingatibuketa, ngasinye singalandzelelwa emuva kuya ekhodini lesivela kuyo. Leyo yincazelo ye-Amazon yensita yayo, hhayi umphumela lociniselekile.

Lifomethi lobhala ngalo umtsetfo libalulekile. Lelinye lemafomethi latiwako yi-EARS, Easy Approach to Requirements Syntax, lelatfutfukiswa ngu-Alistair Mavin nalasebenta nabo e-Rolls-Royce futsi lashicilelwa kwekucala nga-2009. Umcondvo kutsi umtsetfo ngamunye ubhalwa ngekulandzelana lokungashintji: simo, bese kuba sicalo, bese kuba loko sistimu lekufanele ikwente. Linemaphethini lasisekelo langu-5; nanti tibonelo tebhizinisi:

Inzuzo yalelifomethi kutsi umnikati welichubo angalifundza futsi alehlulele ngaphandle kwekwati kuhlela. Umusho munye ngemtsetfo ngamunye, futsi uphendvula ngeyebo, cha noma akukapheleli.

Sibonelo semabhange: inhloso yebhizinisi, hhayi kuhumusha layini ngelayini

Sibonelo lesisha lesicace kakhulu sivela emabhange, lapho imitsetfo yenzuzo, yetimali tensita neyekubhadala sekumashumi eminyaka ihlala ekhodini ye-COBOL kuma-mainframe. I-COBOL lulwimi lwekuhlela loludzala lolusolo lusebentisa emasistimu lamanyenti emabhange nasemshwalensini. Kusimemetelo sangamhla 29 Inyoni 2026, i-Thought Machine, inkampani yebuchwepheshe bemabhange, yatsi isebentisana ne-Amazon kulichubo lelingahumushi likhodi layini ngelayini, kodvwa likhipha inhloso yebhizinisi lehlantekile. Ngekwesimemetelo, lelichubo linetigaba letingu-4:

1. Kukhipha: i-AWS Transform ifundza ema-application lamadala futsi ikhiphe umcondvo webhizinisi uye emininingwaneni lebhalwe nge-EARS.
2. Kuhlanganisa: emakhulu etinguculo letiphindvwako temkhicito munye ahlanganiswa abe yimininingwane lemfisha nalesimanje.
3. Kwakha: ema-ejenti e-AI akha imikhicito yetimali nekuhlolwa kwayo nge-Python.
4. Kucinisekisa: kubuketwa ngebantfu, lapho emacembu ebunjiniyela nemngoti ebhange anelilungelo lelicacile lekuvuma.

Naka timfundvo letimbili: sigaba sekuhlanganisa sikhombisa kutsi sistimu lendala itfwele emakhophi lamanyenti emtsetfo munye, futsi sigaba sekugcina senta ngebantfu ngalokuphelele. Leti timangalo tetinkampani letimbili ngelichubo lato, futsi sibonelo sato libhange, kodvwa indlela iyasebenta kunoma nguyiphi inkampani: khipha, hlanganisa lokuphindziwe, bese uyakha, futsi ungavumi lutfo ngaphambi kwekutsi umnikati wesincumo asayine.

Kungani umnikati welichubo kufanele acinisekise umtsetfo ngamunye

I-AI ifundza loko ikhodi lekwentako, hhayi loko inkampani lekufunako. Sikhala lesiphakatsi kwakho kubili ngiko lapho emaphutsa ekuvuselela enteka khona:

• Likhodi lelingasasebenti: ingakhipha umtsetfo encenyeni lesekuyiminyaka ingasasetjentiswa, iwente ubukeke njengemtsetfo lophilako.
• Liphutsa lelidzala leseliphendvuke umkhuba: sistimu igcwalisa imali ngendlela lengasiyo futsi basebenti bayayilungisa ngesandla. Umtsetfo lokhishiwe utawuhamba nalelophutsa ngaphandle kwekutsi umuntfu atsi liphutsa.
• Umtsetfo ngaphandle kwekhodi: umkhawulo lohlukile lesingatwa ngumcophi wetimali ngesandla ngeke uvele kulokukhishiwe.
• Imitsetfo lephikisanako: tinguculo letimbili tekubala sesaphulelo etindzaweni letimbili letehlukene. Nguyiphi lelungile kusincumo sebhizinisi, hhayi setheknikhi.

Ngako-ke nika umtsetfo ngamunye umnikati waselucheni lwebhizinisi lotawusayina kuwo: umphatsi wekutsengisa wemitsetfo yesaphulelo, umphatsi wetimali wetinhlawulo nekuvuma. Imitsetfo leciniselekile iba sisekelo sencwadzi yetidzingo yesistimu lensha, futsi sichazile kutsi ibhalwa njani kumhlahlandlela wetfu wekubhala tidzingo tesoftware. Khumbula kutsi likhodi netizinda telwati kungaba nedatha yemakhasimende noma yetisebenti, ngako nobe nguliphi lithulusi leliyifundza ngaphandle kweMbuso lidzinga kuhlolwa lesakhuluma ngako encenyeni 3, kanye nemihlahlandlela yeLigunya LaseSaudi Ledatha Nebunhlakanipho Lobentiwe (SDAIA).

Umbono wa-Origami

Nasivuselela sistimu lendala, sitsatsa imitsetfo yayo yebhizinisi njengentfo lebaluleke kakhulu lekufanele idluliselwe, hhayi ikhodi yayo. Sisebentisa emathulusi e-AI kufundza ikhodi nekubhala imitsetfo njengemisho lecacile ngesivinini lebesingenteki ngaphambilini, kodvwa asitsatsi umtsetfo nobe munye njengewekugcina ngaphambi kwekutsi umnikati welichubo awufundze futsi asayine.

Bese sihlola umtsetfo ngamunye lociniselekile ngedatha yangempela yesistimu lendala ngaphambi kwekwakha sistimu lensha luyiphahla, kute sistimu lensha ibale ngendlela inkampani lelindzele ngayo, hhayi ngendlela lithulusi leliyacagela ngayo. Nangabe ufuna kubona kutsi sisebenta njani, buka tinsita tetfu.

Umsebenti waleliviki: umtsetfo munye, ema-invoyisi lamatsatfu

Awudzingi kulindza lithulusi kute ucale. Khetsa umtsetfo munye lovame kuphikiswana ngawo, njengesaphulelo noma inhlawulo yekwephuta, bese ulandzela letinyatselo:

1. Cela umnikati welichubo kutsi abhale umtsetfo ngemusho munye, ngalesakhiwo: Nangabe..., sistimu kufanele...
2. Khipha esistimu ema-invoyisi langempela langu-3 lapho umtsetfo wasebenta khona: linye lelijwayelekile, linye lelisemkhawulweni nalinye lelinemkhawulo lohlukile.
3. Bala i-invoyisi ngayinye ngesandla usebentisa umusho lobhaliwe, bese ucatsanisa umphumela naloko sistimu lekukhiphile.

Nangabe onkhe lamatsatfu ayafana, umtsetfo ubhaliwe. Nangabe linye lehlukile, utfole umtsetfo lofihlekile noma liphutsa lelidzala, futsi loko ngikho kanye lokufanele ukwati ngaphambi kwanobe ngukuphi kuvuselela.

Lokutako

Nyalo unemitsetfo leciniselekile yesistimu yakho. Manje kufika sincumo lesikhulu: ingabe utawuhambisa sistimu injengobe injalo, uguculele ikhodi yayo kuplatfomu yesimanje, uyakhe kabusha uyiphahlise imitsetfo loyikhiphile, noma uyishintje ngemkhicito lowentiwe kakade ugcine idatha kuphela? Encenyeni 5 sicatsanisa letindlela ethebuleni: ngayinye ifaneleka nini, igcinani, ibeka engotini yini, futsi ngubani lokufanele abe yincenye yayo.

Imitfombo

• AWS Transform yema-mainframe: likhasi lelisemtsetfweni — kukhipha imitsetfo yebhizinisi lelandzelelwa ekhodini yekucala, tidzingo letingabuketwa nekuhlolwa kwekulingana.
• Thought Machine ne-AWS: simemetelo sangamhla 29 Inyoni 2026 — tigaba letine: kukhipha nge-EARS, kuhlanganisa, kwakha nekucinisekisa ngebantfu.
• EARS: umhlahlandlela losemtsetfweni wa-Alistair Mavin — lifomethi, emaphethini alo lasihlanu nekushicilelwa kwalo kwekucala nga-2009.
• SDAIA: kuvikela idatha yemuntfu siciniso.