Namuhla, Amamodeli Olimi Olukhulu (LLMs) angafaka isicelo sokuthi aphumelele ukuhlolwa kwe-CPA, kodwa ingabe akulungele ngempela ukuthatha umhlaba wokubalwa kwezimali? Kulesi sihloko, singena kulokho okuzenzakalelayo kwe-AI kubukeka kanjani (nokuthi kungani kungaqondile).

Isingeniso

Akukho ukubalekela iqiniso lokuthi i-AI iyisihloko okukhulunywa ngaso kakhulu ku-inthanethi ngo-2024. I-Chat-GPT, isixhumi esibonakalayo esisekelwe engxoxweni esidumile sokuhlola amakhono e-LLM (Imodeli Yolimi Olukhulu) athuthukiswe yi-OpenAI, yakhishelwa umphakathi. ekuqaleni konyaka.

Dlala ngayo imizuzu embalwa nje, futhi ungaqala ukuqonda ukuthi kungani wonke umuntu nenja yakhe ekhuluma ngalokhu - I-Chat-GPT iyakwazi ukukhombisa ubungcweti obungaphezu kobuntu cishe kuzo zonke izizinda. I-AI ithembisa ngokusobala ukuthi izoguqula kakhulu izindawo eziningi zomsebenzi - kuyilapho inomthelela ezigidini zemisebenzi nemisebenzi.

Ubuhlakani bokwenziwa manje sebusetshenziswa kuzo zonke izizinda zobungcweti esezilungele ukuzenzela - izindawo zomsebenzi ezifana nesoftware, umthetho, ukubalwa kwezimali, ukubonisana, ezezimali nokunye. Ngaphakathi kwezezimali, umsebenzi wokubalwa kwezimali yilowo oza obala njengohlukile ngandlela thize - ikakhulukazi njengoba kubonakala kunomsindo olinganayo ezinhlangothini zombili zengxabano, abameli be-AI kanye nabaphikisayo bobabili banenkulumo-mpikiswano evuthayo mayelana nokuthi yini ezokwenzeka (noma ngeke) kwenzeke.

Ijaji lisacacisa ukuthi lolu shintsho olusheshayo luzofinyelelwa kanjani - futhi yilapho izinkulumo eziningi ezimayelana nezinzuzo ze-ChatGPT ikakhulukazi (kanye ne-AI ngokuvamile) zivame ukudweba umugqa.

Isidingo se-AI ku-Accounting

Emisebenzini yokubala yendabuko, izinkampani zivame ukuthembela ezinqubweni ezenziwa ngesandla, amaphepha amaningi, kanye nemisebenzi ephindaphindwayo ukuze zibambe umsebenzi wazo okhokhelwayo. Le misebenzi iyimisebenzi efana nokufakwa kwedatha, ukucutshungulwa kwe-invoyisi, nokuhlaziywa kwezimali, okubalulekile ekuthathweni kwezinqumo, ukuhlela ukusebenza, nokulawulwa kwengozi.

Kodwa-ke, lezi zinqubo zibandakanya ukuchitha isikhathi (nemali). Izithiyo ezinkulu zomsebenzi we-accounting wezandla yilezi:

Ukufakwa kwedatha mathupha kwethula amathuba aphezulu amaphutha, njengoba abantu bengenza amaphutha lapho befaka idatha ngamavolumu aphezulu. Cabanga ngezinkambu ezifana nezinombolo ze-invoyisi, izinsuku, amanani edola - ukuthola noma yikuphi kwalokhu okungalungile kunemiphumela emikhulu.

Kudla isikhathi, kudinga amahora amaningi okusebenza ukuze kuhlanganiswe ama-akhawunti, kwenziwe imibiko, futhi kwenziwe ukuhlaziya kwezezimali.

Kunzima ekuxhumaneni okuvumelanayo. Ingabe uke wabhekana nezimo ezifana nalezi ezingezansi?

a. Ukugunyaza akwenzeki kuze kube yilapho uthola iklayenti kanye ne-CPA ocingweni

b. Izinto zolayini azixazululeki kuze kube yilapho iklayenti lihlela umhlangano nethimba lakho elifaka idatha ye-invoyisi kanye nokuphathwa kwedokhumenti.

Konke lokhu kuholela ekubambezelekeni kokuvalwa kwanyanga zonke kwamakhasimende, ukukhokhelwa kwabathengisi sekwephuzile, ukungahleleki kwezindleko, kanye nobunzima ekugcineni ubuqotho bezezimali.

I-AI ye-accounting ayidingi ukusho ukulungiswa okuphelele

Izinkinga ezibalwe ngenhla zibhalwe kahle - futhi lapho ebuzwa, amaqembu amaningi e-accounting azovuma ukuthi ukwethula i-AI kuzobasiza nakanjani. Ubuchwepheshe obufana nokufunda komshini nokucutshungulwa kolimi lwemvelo bunamandla okuguqula umsebenzi wokubala ngendlela ejule kakhulu - inqobo nje uma busetshenziswa futhi buhlanganiswe ngendlela efanele.

Kodwa-ke, lokhu kuvame ukuholela abaningi esiphethweni sokuthi i-AI-based automation akuyona eyabo - kubonakala kunzima, kudla isikhathi futhi kuyabiza ukukusebenzisa.

Iqiniso, nokho, ngeke lihluke kakhulu - namuhla kungenzeka ukuthi uqalise ngokusebenzisa i-AI ngenqubo yakho yokubala ngemizuzu. Futhi ungakwazi ukufeza lokhu ngaphandle kokuyekethisa ukwethembeka kwenqubo yakho yamanje, ukuphepha nokusebenza kahle.

Beka eceleni i-AI nama-LLM akhiqizayo umzuzwana owodwa - iqiniso liwukuthi ngisho ne-automation yezinga lokungena ingasiza kakhulu ekuxazululeni lezi zinkinga. Ngisho ne-OCR ethobekile - osekungamashumi eminyaka ikhona - inciphisa isikhathi esithathwa ukucubungula i-invoyisi okungenani ngo-60%, konga amaqembu amabhuku ezimali izinsuku eziningi njalo ngenyanga. Nokho ukwamukelwa kwalobu buchwepheshe akukandile.

Amacala angase asetshenziswe ku-AI ngaphakathi kwenqubo ye-Accounting

Ngakho-ke kufanele uyihlanganise kanjani ngempela i-AI ohlelweni lwakho lokubala? Uqala kuphi?

Indawo yokuqala okufanele uqale ngayo ukubheka ukuthi iyiphi ingxenye yenqubo ethatha isikhathi esiningi. Izingqinamba ezijwayelekile ezibikwa ngamathimba amabhuku ezimali ziyimisebenzi efana nale:

1. Ukubhala ama-invoyisi
2. I-General Ledger (GL) yokwenza imephu
3. Ukuqinisekiswa Kwemininingwane Yenkokhelo (ukuze uhlole ukukhwabanisa)
4. Ukutholwa Kwempinda

Kunendikimba eyisisekelo ecace kakhulu lapha - ukufaka idatha okwenziwa ngesandla kanye nokuqinisekisa yikho okubangela ukuthi le misebenzi ibe isidina futhi idle isikhathi.

Lo mfanekiso wenhlolovo ongenhla (ovela embikweni we-Automation Trends 2022) wembula okuningi - cishe u-70% wabantu abakazenzi ngokuzenzakalelayo izinkinga ezibacindezela kakhulu enqubweni yabo yokubala. Imisebenzi ebalwe ngenhla yonke ingeyemanyuwali - othile udinga ukubheka idatha yangempela ku-invoyisi futhi aqinisekise ukuthi ilungile, ngaphambi kokuqhubeka.

Kanjalo, ukwenza ngokuzenzakalelayo le misebenzi kungase kuzwakale kungaphezu kwamandla, njengoba manje usuthemba umshini ukuthi uzoba nezinga elifanayo lokuqonda njengelomuntu (oqeqeshiwe).

Izindaba ezinhle? I-AI ingaqeqeshwa ngokulinganayo futhi! Singena ngokujulile kwezinye izimo zokusetshenziswa kwalokhu, ngezansi.

1. Ukubhalwa kwekhodi kwe-invoyisi kanye ne-General Ledger (GL) yokwenza imephu ye-akhawunti

Mhlawumbe omunye wemisebenzi enzima kakhulu ukuwenza ngokuzenzakalelayo ukwaba ama-invoyisi namarisidi esigabeni esifanele kanye nekhodi ye-GL ngaphakathi kwesistimu yakho yokubala. Kungani lokhu kuwubuqili ngokukhethekile?

Kuvame ukuba namakhodi amaningi e-GL asebenza ezindlekweni ezifanayo, ahlukaniswa ngezinto zomugqa/amakhodi omkhiqizo ngamunye. Ukunikezwa kwalawa makhodi e-GL ngokuvamile kwenziwa ngesandla, futhi kufanele kwenziwe ngokubonisana namaqembu ebhizinisi kanye ne-CFO.

Ukunikeza ikhodi ye-GL ku-invoyisi ngezinye izikhathi kuwukuziphendulela - isibonelo, kuyilapho ama-invoyisi okuthengisa avamile angase anikezwe "Ezokuthengisa" eshadini lakho lama-akhawunti, ngezinye izikhathi ifomethi ye-invoyisi efanayo ncamashi igcina isetshenziselwa osonkontileka nabangebona abasebenzi. Lokhu kungaholela ekutheni izindleko zenkontileka zimakwe ngokungalungile njengokuthi “Okuthengiswayo” ngamathuluzi ayisisekelo e-automation.

I-AI ingasiza kanjani lapha?

Yenza ngokuzenzakalelayo ikhodi ye-invoyisi ngokusekelwe ekucutshungulweni kwe-LLM - lapha, i-AI ikutshela ngokuyisisekelo ukuthi iyiphi i-GL le invoyisi okufanele ihlukaniswe ngayo, futhi lokhu kungalungiselelwa ukunikeza iziphakamiso eziningi ezingaba ezifanele. Lokhu kwenza umsebenzi womsebenzisi ube lula.

Funda futhi ubambe ngekhanda okokufaka komsebenzisi - uma umsebenzisi esekhethe ikhodi ye-GL, isistimu ingakhumbula okukhethiwe futhi iyenze ngokuzenzakalelayo ngesikhathi esilandelayo kumthengisi ofanayo.

2. Ukutholwa kokukhwabanisa nokuphatha amaphutha

Omunye umsebenzi obalulekile ithimba labacwaningi mabhuku elinawo wukubamba amaphutha ngaphambi kokuthi enzeke. Kungase kube bucayi njengemininingwane yokukhokha okungalungile kanye nokukhwabanisa kwe-invoyisi, noma kungase kube lula njenge-invoyisi eyimpinda.

Ngaphandle kokungabaza, lezi zinkinga zigwenywa kangcono ngaphambi kokuba zenzeke. Iningi lezinhlangano liyaphikelela ekwenzeni le manuwali yenqubo. Kodwa-ke, ukuhlola i-invoyisi ngayinye kwenza izinto zibe nzima ngoba:

Inikeza iphuzu elilodwa lokwehluleka (kanye nebhodlela) lenqubo - kuyilapho kuhle ukuthi isisebenzi sihlole zonke izindleko ukuze sithole amaphutha, ngezinye izikhathi izinto zingangena emifantwini.

Iqinisekisa ukuthi kuphela umuntu onomongo omningi ekufakweni kokubalwa kwezimali (i-CFO/inhloko ye-accounting) ongenza izilungiso, futhi akekho omunye. Lonke ulwazi nomongo kunabantu abambalwa kuphela, futhi akusakazeki kuyo yonke inhlangano.

I-AI ingasiza kanjani lapha?

Ukutholwa kwempinda ehlakaniphile/ulwazi olungalungile - Ukuhlolwa kwempinda yefayela eyisisekelo kuqinisekisa kuphela uma amafayela amabili afana. Ngokuhlola okuphindwe kabili kwe-AI okuthuthukisiwe, ungaqhubeka nesinyathelo esisodwa - ubheka ukuthi okuqukethwe kwamafayela amabili ahlukene kufana ngokusolisayo yini.

Ukuqinisekiswa kwedatha okuningi kudatha ye-invoyisi - Ukufunda ngokuzenzakalelayo idatha ye-invoyisi akusizi uma othile kufanele angene ngemvume futhi ayiqinisekise noma kunjalo. Amathuluzi e-AI athuthukile manje angakwazi ukwenza ukuqinisekiswa kwedatha ukuze kuqinisekiswe ukuhlolwa kwenhlanzeko (isibonelo, uma inombolo ye-akhawunti yasebhange entsha ku-invoyisi ingafani nejwayelekile yomthengisi, uzokwaziswa!)

3. Ukufunda izenzo ezilula eziphindaphindekayo

Buza noma ubani ukuthi NGEMPELA bafuna i-AI yenzeni, futhi lena impendulo ephuma phezulu - abantu abaningi banomuzwa wokuthi inani langempela le-AI yilapho ingakwazi ukufunda amaphethini abo futhi yonge isikhathi sabo.

Isibonelo, kunemisebenzi emincane eminingi eyenziwa ngendlela efanayo, ezinhlotsheni eziningi zama-invoyisi/amarisidi. Ezinye izibonelo:

Ukunikeza i-invoyisi esigabeni esilungile/isigaba/iphrojekthi ku-ERP yakho

Ukushintsha imephu ye-GL yento eyodwa yomugqa we-invoyisi

Ukuthumela i-invoyisi yomthengisi othile ukuze ivunyelwe kumuntu ofanayo, ngaso sonke isikhathi

I-AI ingasiza kanjani lapha?

Isinyathelo sokuqala ukukhomba izinyathelo zenqubo yokubalwa kwezimali ezifanele kahle ukufunda kabusha okuphindaphindwayo (okungukuthi, imisebenzi oqhubeka uyenza nsuku zonke, engagcina ibanjwe ngekhanda i-AI futhi izenzele ngokuzenzakalela u-90% wesikhathi).

Izibonelo ezinhle zalokhu yilezi:

Umsebenzi wekhodi ye-GL - Umqondo lapha ulula: uma uhlelo lokusebenza lunikeza ikhodi ye-GL efanelekile ku-invoyisi, kuhle! Uma kungenjalo, uzishintshela wena, futhi i-AI ikhumbula lolu shintsho ngokuzayo. Ngenxa yalokho, umsebenzi ozokwenziwa ngekhodi ye-GL ozenzakalelayo uqhubeka ubangcono ngakho konke ukuchofoza okwenzayo.

Isigaba/Ikilasi/Iphrojekthi – Uma i-invoyisi ethile yomthengisi ingakwazi ukuhlelwa ngokuzenzakalelayo esigabeni esifanele, i-AI ingafunda amaphethini ekukhetheni kwakho (ngokwesibonelo, ingabe uhlale uhlukanisa amarisidi e-Uber ngokuthi “Izindleko Zephrojekthi” esikhundleni sokuthi “Ukuhamba”?). Ngokuhamba kwesikhathi, lokhu kuba umthetho omisiwe ngaphakathi kwenkundla yakho, futhi usetshenziswa ngokuzenzakalelayo.

I-Nanonets ingakusiza kanjani ukuthi usebenzise i-AI kuNqubo yakho yokubala

Izibonelo ezingenhla cishe ziyisihloko se-iceberg - kuningi okungaphezu kwalokho i-AI engakwenza ngenqubo yakho yokubala okunqunyelwe kuphela ukuthi ukwazi ukujula kangakanani enqubweni yokuzenzakalela nokufunda ngomshini.

Ngenhlanhla, namuhla akudingekile ukuba ube nolwazi lobuchwepheshe ukuze uqale ukusebenzisa amakhono e-AI ohlelweni lwakho lokubala - akhona amathuluzi akuvumela ukuthi uqalise cishe ngokushesha.

Isibonelo, i-Nanonets iyinkundla ye-AI engaguqula inqubo yakho yamanje yokubala, futhi yengeze lezo zici ezibalulekile ze-AI ekuhambeni kwakho komsebenzi. Ingenza konke okuboniswe ngenhla - nokunye okuningi.

Kulula ukusebenzisa nokho okuyinkimbinkimbi ngamakhono ako, lesi indawo ekahle yokuqala yalabo abafuna ukukhulisa inqubo yabo yokubala futhi bakale umthwalo wabo ngempumelelo. Thintana namuhla ukuze uthole umboniso wamahhala wokuthi le nkundla ye-AI ingakwenzelani umsebenzi wakho wokubala.

• I-SEO Powered Content & PR Distribution. Khuliswa Namuhla.
• I-PlatoData.Network Vertical Generative Ai. Zinike Amandla. Finyelela Lapha.
• I-PlatoAiStream. I-Web3 Intelligence. Ulwazi Lukhulisiwe. Finyelela Lapha.
• I-PlatoESG. Ikhabhoni, I-CleanTech, Amandla, Environment, Ilanga, Ukuphathwa Kwemfucuza. Finyelela Lapha.
• I-PlatoHealth. I-Biotech kanye ne-Clinical Trials Intelligence. Finyelela Lapha.
• Source: https://nanonets.com/blog/ai-automation-for-accounting-2/