中國人民銀行《金融(rong)科技(ji)發展規劃（2022~2025年》指出：運用機器人流(liu)(liu)程(cheng)自(zi)動化（RPA）、自(zi)然(ran)語言處理(li)（NLP）、智(zhi)能(neng)字符識別（ICR）等(deng)智(zhi)能(neng)技(ji)術開(kai)展(zhan)端到端數字化流(liu)(liu)程(cheng)重構，打(da)通部門間業(ye)務阻隔與(yu)流(liu)(liu)程(cheng)斷點，實現跨角色、跨時序的業(ye)務靈活(huo)定制與(yu)編排，打(da)造環節無縫銜接、信息實時交(jiao)互、資源協同高效(xiao)的業(ye)務處理(li)模(mo)式，更好支撐(cheng)數字化業(ye)務發(fa)展(zhan)。

人(ren)保財險科技(ji)運營(ying)部自2019年(nian)(nian)以來探索引入(ru)RPA技(ji)術，打(da)(da)造數字員(yuan)工，將(jiang)大量重復且規則明(ming)確的(de)業(ye)務操(cao)作(zuo)由機(ji)器人(ren)自動(dong)執(zhi)行替(ti)代人(ren)工操(cao)作(zuo)，節省(sheng)員(yuan)工在機(ji)械、重復、耗(hao)時(shi)工作(zuo)上的(de)工作(zuo)量和時(shi)間，使員(yuan)工可以從事(shi)更有(you)(you)價(jia)值和創(chuang)造性的(de)工作(zuo)，有(you)(you)效(xiao)提升運營(ying)效(xiao)能和人(ren)力績(ji)效(xiao)。經(jing)過兩年(nian)(nian)左右的(de)不(bu)斷嘗試(shi)并積累(lei)經(jing)驗(yan)，2022年(nian)(nian)，科技(ji)運營(ying)部將(jiang)數字員(yuan)工平臺(tai)項目納入(ru)人(ren)保財險“三化(hua)(hua)”建設(she)首批項目清(qing)單，通過建章立制、組建CoE（Center of Excellence，一(yi)般(ban)稱(cheng)卓越中(zhong)心，負(fu)責(ze)統(tong)籌推動(dong)RPA應用(yong)）、深挖場景、打(da)(da)造平臺(tai)、強化(hua)(hua)管控(kong)、技(ji)術創(chuang)新(xin)等(deng)舉措，自上而下有(you)(you)組織(zhi)地推動(dong)RPA系(xi)統(tong)化(hua)(hua)、規范化(hua)(hua)、規模化(hua)(hua)應用(yong)，促進降本增效(xiao)，助力公司(si)運營(ying)自動(dong)化(hua)(hua)和數字化(hua)(hua)轉型(xing)。

截至2022年11月(yue)底，數(shu)字(zi)員工(gong)(gong)(gong)平臺已在人保財險總分公(gong)(gong)司(si)上線612個RPA流程，打造(zao)699名數(shu)字(zi)員工(gong)(gong)(gong)，涵蓋總部(bu)及33家省級分公(gong)(gong)司(si)。2022年以來，數(shu)字(zi)員工(gong)(gong)(gong)累計運(yun)行時長超(chao)(chao)過(guo)84萬小時，節約人工(gong)(gong)(gong)工(gong)(gong)(gong)作(zuo)量超(chao)(chao)過(guo)356人年，有(you)效(xiao)提(ti)升了業(ye)務運(yun)營效(xiao)率和人力效(xiao)能，實現(xian)降本增效(xiao)。

廣泛應用(yong)的車險核保機器人。在(zai)車(che)險(xian)業(ye)(ye)務(wu)中，存在(zai)大量規則清晰且重復性高(gao)的(de)核(he)(he)保工(gong)(gong)作(zuo)(zuo)(zuo)，核(he)(he)保處(chu)理(li)(li)時需要(yao)操(cao)作(zuo)(zuo)(zuo)多個系統，人(ren)(ren)工(gong)(gong)處(chu)理(li)(li)的(de)工(gong)(gong)作(zuo)(zuo)(zuo)量、復雜(za)度、易錯程度都較高(gao)，處(chu)理(li)(li)的(de)時效性也不能完(wan)全滿(man)足業(ye)(ye)務(wu)需要(yao)。針(zhen)對這一場景，開(kai)發車(che)險(xian)核(he)(he)保機器(qi)人(ren)(ren)輔助人(ren)(ren)工(gong)(gong)處(chu)理(li)(li)核(he)(he)保工(gong)(gong)作(zuo)(zuo)(zuo)，通過RPA融合(he)OCR技(ji)術(shu)，提升(sheng)信息審核(he)(he)速度及準確性，輔助人(ren)(ren)工(gong)(gong)快(kuai)速完(wan)成核(he)(he)保操(cao)作(zuo)(zuo)(zuo)，效率提升(sheng)600%。采用總分協同模式對數字(zi)員工(gong)(gong)進行集中管理(li)(li)，目前車(che)險(xian)核(he)(he)保機器(qi)人(ren)(ren)已在(zai)28家省級(ji)分公(gong)司(si)上(shang)線(xian)(xian)，2022年以(yi)來已累計節約37人(ren)(ren)年工(gong)(gong)作(zuo)(zuo)(zuo)量。其(qi)中，人(ren)(ren)保財險(xian)福建分公(gong)司(si)共(gong)上(shang)線(xian)(xian)9個車(che)險(xian)核(he)(he)保機器(qi)人(ren)(ren)，平均每天處(chu)理(li)(li)300單業(ye)(ye)務(wu)，替(ti)代7名人(ren)(ren)工(gong)(gong)。

集中部署的理賠(pei)集審照片分揀機器人(ren)。在(zai)理(li)(li)(li)賠(pei)業(ye)務(wu)中，在(zai)線理(li)(li)(li)賠(pei)坐(zuo)席(xi)需要通過(guo)集中審核平(ping)臺(tai)對(dui)各個渠道上傳的(de)海量照(zhao)片(pian)進(jin)行(xing)(xing)人(ren)(ren)(ren)工審核和分揀，工作量大，耗費人(ren)(ren)(ren)力成(cheng)本高。針對(dui)這一場(chang)景，開發理(li)(li)(li)賠(pei)集審照(zhao)片(pian)分揀機器(qi)人(ren)(ren)(ren)，通過(guo)RPA融合OCR技(ji)術(shu)對(dui)照(zhao)片(pian)關(guan)鍵信(xin)息進(jin)行(xing)(xing)自動提(ti)(ti)取和識別，將(jiang)識別后的(de)照(zhao)片(pian)自動分揀至(zhi)對(dui)應目錄，大幅提(ti)(ti)升線上理(li)(li)(li)賠(pei)處(chu)理(li)(li)(li)時效。理(li)(li)(li)賠(pei)集審照(zhao)片(pian)分揀機器(qi)人(ren)(ren)(ren)自2021年年底(di)上線以來，采用總部(bu)集中部(bu)署的(de)方式，共部(bu)署15名數(shu)字員(yuan)工，累(lei)計運行(xing)(xing)22019小時，節約人(ren)(ren)(ren)力1633人(ren)(ren)(ren)天，有效提(ti)(ti)高了分揀效率(lv)和準確率(lv)，在(zai)為理(li)(li)(li)賠(pei)坐(zuo)席(xi)減負的(de)同時，將(jiang)數(shu)字員(yuan)工的(de)處(chu)理(li)(li)(li)結(jie)果(guo)記(ji)錄在(zai)表格中，每(mei)日交由指定人(ren)(ren)(ren)員(yuan)對(dui)結(jie)果(guo)進(jin)行(xing)(xing)二次檢查(cha)，保證流程運行(xing)(xing)平(ping)穩且風險可控。

處理短期大批(pi)量業務(wu)的農險業務(wu)機器人。農(nong)險(xian)(xian)(xian)(xian)業(ye)(ye)務(wu)具有受政(zheng)(zheng)策(ce)(ce)約束(shu)較大(da)、集中(zhong)(zhong)出單(dan)占比(bi)高、季(ji)節性較強(qiang)等特點。在一些特殊(shu)的(de)(de)應(ying)用場景，如財政(zheng)(zheng)部(bu)(bu)下(xia)(xia)(xia)發《中(zhong)(zhong)央(yang)財政(zheng)(zheng)農(nong)業(ye)(ye)保(bao)險(xian)(xian)(xian)(xian)保(bao)費補貼管(guan)理辦法》要求(qiu)調(diao)整政(zheng)(zheng)策(ce)(ce)性農(nong)險(xian)(xian)(xian)(xian)各險(xian)(xian)(xian)(xian)種的(de)(de)各級財政(zheng)(zheng)補貼比(bi)例、農(nong)險(xian)(xian)(xian)(xian)集中(zhong)(zhong)承(cheng)保(bao)錄單(dan)時進行(xing)(xing)海量(liang)影像上(shang)(shang)傳(chuan)(chuan)、農(nong)險(xian)(xian)(xian)(xian)理賠大(da)客戶(hu)案(an)件(jian)抽檢要求(qiu)下(xia)(xia)(xia)載系統案(an)件(jian)清單(dan)及影像資(zi)料時，業(ye)(ye)務(wu)量(liang)巨大(da)且對時效(xiao)性及數(shu)據質量(liang)要求(qiu)都很高，分(fen)公司往往需(xu)要抽調(diao)大(da)量(liang)業(ye)(ye)務(wu)人(ren)(ren)(ren)員(yuan)集中(zhong)(zhong)在數(shu)天(tian)或(huo)(huo)數(shu)十(shi)天(tian)內加(jia)班加(jia)點完成上(shang)(shang)萬件(jian)保(bao)單(dan)的(de)(de)錄單(dan)、批改或(huo)(huo)資(zi)料下(xia)(xia)(xia)載操作(zuo)(zuo)。為提高業(ye)(ye)務(wu)處理效(xiao)率，開(kai)發上(shang)(shang)線(xian)農(nong)險(xian)(xian)(xian)(xian)政(zheng)(zheng)策(ce)(ce)性系數(shu)批改機器(qi)人(ren)(ren)(ren)、農(nong)險(xian)(xian)(xian)(xian)影像上(shang)(shang)傳(chuan)(chuan)機器(qi)人(ren)(ren)(ren)、農(nong)險(xian)(xian)(xian)(xian)理賠資(zi)料下(xia)(xia)(xia)載機器(qi)人(ren)(ren)(ren)等農(nong)險(xian)(xian)(xian)(xian)業(ye)(ye)務(wu)機器(qi)人(ren)(ren)(ren)。根據業(ye)(ye)務(wu)需(xu)要，快速部(bu)(bu)署(shu)數(shu)十(shi)名數(shu)字員(yuan)工(gong)，完成短期(qi)大(da)批量(liang)任務(wu)后再將機器(qi)人(ren)(ren)(ren)許(xu)可資(zi)源釋(shi)放，極大(da)地(di)發揮了(le)數(shu)字員(yuan)工(gong)快速部(bu)(bu)署(shu)、靈(ling)活適配業(ye)(ye)務(wu)的(de)(de)優勢，減(jian)輕了(le)基層一線(xian)的(de)(de)工(gong)作(zuo)(zuo)負擔。2022年7月，人(ren)(ren)(ren)保(bao)財險(xian)(xian)(xian)(xian)河南(nan)分(fen)公司緊(jin)急開(kai)發上(shang)(shang)線(xian)并部(bu)(bu)署(shu)15個農(nong)險(xian)(xian)(xian)(xian)政(zheng)(zheng)策(ce)(ce)性系數(shu)批改機器(qi)人(ren)(ren)(ren)，共(gong)運行(xing)(xing)2個月，完成23890單(dan)批改工(gong)作(zuo)(zuo)，節約人(ren)(ren)(ren)力(li)498人(ren)(ren)(ren)天(tian)。

開創數字員工服務化(hua)模(mo)式的運維機器人。人保(bao)(bao)財險(xian)湖(hu)南(nan)(nan)(nan)(nan)分公司科(ke)技運(yun)(yun)(yun)營(ying)(ying)部率先提出(chu)“數(shu)(shu)字員(yuan)(yuan)工(gong)(gong)服(fu)務化(hua)”理念(nian)并(bing)實現落地，自(zi)(zi)主開發(fa)“小南(nan)(nan)(nan)(nan)同(tong)學(xue)”智能運(yun)(yun)(yun)維(wei)應(ying)(ying)用。一方(fang)面依(yi)托(tuo)企業(ye)微(wei)信工(gong)(gong)作臺作為入口，接收用戶以(yi)自(zi)(zi)然語言方(fang)式(shi)下達的(de)(de)任務命令(ling)并(bing)轉化(hua)為數(shu)(shu)字員(yuan)(yuan)工(gong)(gong)可(ke)以(yi)理解的(de)(de)任務消(xiao)息，另一方(fang)面將各(ge)類運(yun)(yun)(yun)維(wei)數(shu)(shu)字員(yuan)(yuan)工(gong)(gong)封(feng)裝成標準化(hua)服(fu)務以(yi)便調用。各(ge)數(shu)(shu)字員(yuan)(yuan)工(gong)(gong)監(jian)聽消(xiao)息，當(dang)監(jian)聽到歸屬于自(zi)(zi)身的(de)(de)任務消(xiao)息時則自(zi)(zi)動(dong)執(zhi)行(xing)，并(bing)將執(zhi)行(xing)結果(guo)返(fan)回“小南(nan)(nan)(nan)(nan)同(tong)學(xue)”前端(duan)并(bing)反饋用戶。“小南(nan)(nan)(nan)(nan)同(tong)學(xue)”將日常(chang)運(yun)(yun)(yun)維(wei)中的(de)(de)重(zhong)復工(gong)(gong)作交給數(shu)(shu)字員(yuan)(yuan)工(gong)(gong)自(zi)(zi)動(dong)解決，如車險(xian)投保(bao)(bao)預確認、重(zhong)新(xin)(xin)轉保(bao)(bao)單、電子保(bao)(bao)單重(zhong)發(fa)等(deng)，實現運(yun)(yun)(yun)維(wei)工(gong)(gong)作的(de)(de)智能化(hua)自(zi)(zi)執(zhi)行(xing)，并(bing)將數(shu)(shu)字員(yuan)(yuan)工(gong)(gong)的(de)(de)調用無縫(feng)融入運(yun)(yun)(yun)維(wei)工(gong)(gong)作場景(jing)，支持數(shu)(shu)字員(yuan)(yuan)工(gong)(gong)的(de)(de)自(zi)(zi)定義(yi)擴展和快(kuai)速接入，實現“數(shu)(shu)字員(yuan)(yuan)工(gong)(gong)服(fu)務化(hua)”模式(shi)。人保(bao)(bao)財險(xian)科(ke)技運(yun)(yun)(yun)營(ying)(ying)部基(ji)于湖(hu)南(nan)(nan)(nan)(nan)分公司創新(xin)(xin)探(tan)索，打造統頒(ban)“人保(bao)(bao)小智”智能運(yun)(yun)(yun)維(wei)應(ying)(ying)用并(bing)進行(xing)全國推(tui)廣(guang)。

數(shu)字員工具有廣闊的應(ying)(ying)用(yong)(yong)前景。下一步，人(ren)(ren)保財險科技(ji)運(yun)營部將(jiang)持續深(shen)(shen)化數(shu)字員工建(jian)設和應(ying)(ying)用(yong)(yong)，協同(tong)(tong)業(ye)務部門及分公司(si)進(jin)(jin)一步挖掘(jue)、擴大應(ying)(ying)用(yong)(yong)場(chang)景，優化、完善(shan)數(shu)字員工運(yun)營管理(li)平臺(tai)支撐能力(li)，提升(sheng)數(shu)字員工效能以及運(yun)行穩(wen)定性，加(jia)強(qiang)總分協同(tong)(tong)、技(ji)術(shu)培訓(xun)和流程推廣，深(shen)(shen)化RPA與AI等(deng)技(ji)術(shu)融(rong)合(he)創(chuang)新，積極與同(tong)(tong)業(ye)交(jiao)流學習(xi)先(xian)進(jin)(jin)經驗(yan)，探索應(ying)(ying)用(yong)(yong)RPM（機器人(ren)(ren)流程挖掘(jue)）技(ji)術(shu)發(fa)掘(jue)更多(duo)應(ying)(ying)用(yong)(yong)場(chang)景，進(jin)(jin)一步促進(jin)(jin)降(jiang)本增效，助(zhu)力(li)運(yun)營向自(zi)動化和智能化轉型。