中(zhong)國(guo)人民銀行《金融科技發展(zhan)規(gui)劃（2022~2025年》指(zhi)出：運用機器人(ren)流(liu)(liu)程(cheng)自動化(hua)（RPA）、自然語言處理（NLP）、智能字(zi)符識(shi)別（ICR）等智能技(ji)術開(kai)展端(duan)到端(duan)數(shu)字(zi)化(hua)流(liu)(liu)程(cheng)重構，打通部門間業(ye)務阻隔與流(liu)(liu)程(cheng)斷點，實現跨角色、跨時(shi)(shi)序的業(ye)務靈活定制與編排，打造(zao)環(huan)節無縫(feng)銜(xian)接、信息實時(shi)(shi)交互(hu)、資源協同高(gao)效的業(ye)務處理模(mo)式，更好支撐數(shu)字(zi)化(hua)業(ye)務發展。

人保財(cai)險(xian)科技運營(ying)部(bu)(bu)自(zi)2019年以來探索引入RPA技術(shu)，打造數(shu)字員工(gong)(gong)，將(jiang)大量重復且規則明(ming)確的業(ye)務操(cao)作由機器人自(zi)動執行替代人工(gong)(gong)操(cao)作，節(jie)省(sheng)員工(gong)(gong)在機械、重復、耗時工(gong)(gong)作上的工(gong)(gong)作量和(he)(he)時間，使員工(gong)(gong)可以從事更有(you)價值和(he)(he)創(chuang)造性的工(gong)(gong)作，有(you)效提升運營(ying)效能和(he)(he)人力績效。經過兩年左右的不斷嘗試(shi)并積累經驗，2022年，科技運營(ying)部(bu)(bu)將(jiang)數(shu)字員工(gong)(gong)平(ping)臺項目納入人保財(cai)險(xian)“三(san)化(hua)(hua)(hua)”建(jian)設首批項目清單，通(tong)過建(jian)章立制、組建(jian)CoE（Center of Excellence，一般稱(cheng)卓越(yue)中(zhong)心，負責統籌推動RPA應用）、深(shen)挖場景、打造平(ping)臺、強化(hua)(hua)(hua)管控、技術(shu)創(chuang)新(xin)等舉(ju)措(cuo)，自(zi)上而下有(you)組織(zhi)地(di)推動RPA系(xi)統化(hua)(hua)(hua)、規范化(hua)(hua)(hua)、規模(mo)化(hua)(hua)(hua)應用，促進(jin)降本(ben)增(zeng)效，助力公司運營(ying)自(zi)動化(hua)(hua)(hua)和(he)(he)數(shu)字化(hua)(hua)(hua)轉型。

截至(zhi)2022年(nian)11月底(di)，數字員工(gong)平臺已在(zai)人(ren)保財(cai)險(xian)總分公司(si)上線612個RPA流程，打造(zao)699名(ming)數字員工(gong)，涵蓋總部及33家省級分公司(si)。2022年(nian)以來，數字員工(gong)累計運(yun)行時(shi)長超過84萬小時(shi)，節約(yue)人(ren)工(gong)工(gong)作量超過356人(ren)年(nian)，有(you)效(xiao)提(ti)升了業務運(yun)營效(xiao)率和人(ren)力(li)效(xiao)能，實現降本增效(xiao)。

廣泛應用的車險核保機器人。在車(che)(che)險(xian)業務(wu)中(zhong)，存在大量(liang)規則清晰且重(zhong)復性高的核(he)保(bao)(bao)(bao)工(gong)(gong)(gong)(gong)(gong)(gong)作，核(he)保(bao)(bao)(bao)處(chu)理(li)時需要(yao)操(cao)作多個(ge)系統，人(ren)(ren)工(gong)(gong)(gong)(gong)(gong)(gong)處(chu)理(li)的工(gong)(gong)(gong)(gong)(gong)(gong)作量(liang)、復雜度(du)、易(yi)錯程度(du)都較高，處(chu)理(li)的時效性也不能完(wan)(wan)全(quan)滿足業務(wu)需要(yao)。針對(dui)這一場景，開發車(che)(che)險(xian)核(he)保(bao)(bao)(bao)機器人(ren)(ren)輔助(zhu)人(ren)(ren)工(gong)(gong)(gong)(gong)(gong)(gong)處(chu)理(li)核(he)保(bao)(bao)(bao)工(gong)(gong)(gong)(gong)(gong)(gong)作，通過(guo)RPA融合OCR技術，提升信息審(shen)核(he)速度(du)及準確性，輔助(zhu)人(ren)(ren)工(gong)(gong)(gong)(gong)(gong)(gong)快速完(wan)(wan)成核(he)保(bao)(bao)(bao)操(cao)作，效率提升600%。采用總分協同模式(shi)對(dui)數(shu)字員(yuan)工(gong)(gong)(gong)(gong)(gong)(gong)進行集中(zhong)管理(li)，目前車(che)(che)險(xian)核(he)保(bao)(bao)(bao)機器人(ren)(ren)已在28家省級(ji)分公司(si)上線，2022年(nian)以來(lai)已累(lei)計節約37人(ren)(ren)年(nian)工(gong)(gong)(gong)(gong)(gong)(gong)作量(liang)。其中(zhong)，人(ren)(ren)保(bao)(bao)(bao)財險(xian)福建分公司(si)共上線9個(ge)車(che)(che)險(xian)核(he)保(bao)(bao)(bao)機器人(ren)(ren)，平(ping)均每天處(chu)理(li)300單業務(wu)，替(ti)代7名人(ren)(ren)工(gong)(gong)(gong)(gong)(gong)(gong)。

集(ji)中部署的理賠集(ji)審照片分揀機器(qi)人。在理(li)賠(pei)業務中(zhong)(zhong)，在線理(li)賠(pei)坐(zuo)席需要通過集中(zhong)(zhong)審(shen)(shen)核平(ping)臺對各個(ge)渠道上(shang)(shang)傳的(de)(de)海(hai)量照(zhao)(zhao)(zhao)片(pian)進(jin)(jin)(jin)行(xing)人(ren)(ren)工審(shen)(shen)核和分揀，工作量大，耗費(fei)人(ren)(ren)力成(cheng)本高。針對這(zhe)一場(chang)景，開發(fa)理(li)賠(pei)集審(shen)(shen)照(zhao)(zhao)(zhao)片(pian)分揀機(ji)器(qi)人(ren)(ren)，通過RPA融合OCR技術對照(zhao)(zhao)(zhao)片(pian)關鍵信息進(jin)(jin)(jin)行(xing)自(zi)動提(ti)取和識(shi)別，將識(shi)別后的(de)(de)照(zhao)(zhao)(zhao)片(pian)自(zi)動分揀至對應目錄，大幅(fu)提(ti)升線上(shang)(shang)理(li)賠(pei)處理(li)時效(xiao)。理(li)賠(pei)集審(shen)(shen)照(zhao)(zhao)(zhao)片(pian)分揀機(ji)器(qi)人(ren)(ren)自(zi)2021年(nian)年(nian)底上(shang)(shang)線以來，采用(yong)總部(bu)集中(zhong)(zhong)部(bu)署的(de)(de)方式，共部(bu)署15名數字員工，累計(ji)運行(xing)22019小時，節(jie)約人(ren)(ren)力1633人(ren)(ren)天，有效(xiao)提(ti)高了分揀效(xiao)率(lv)和準確率(lv)，在為(wei)理(li)賠(pei)坐(zuo)席減負的(de)(de)同時，將數字員工的(de)(de)處理(li)結(jie)果記錄在表(biao)格中(zhong)(zhong)，每日交由指定人(ren)(ren)員對結(jie)果進(jin)(jin)(jin)行(xing)二次檢(jian)查，保證流程運行(xing)平(ping)穩且(qie)風險可(ke)控。

處理(li)短期大(da)批量(liang)業(ye)務的農險業(ye)務機(ji)器(qi)人。農(nong)(nong)(nong)險(xian)(xian)(xian)業(ye)務(wu)具有受(shou)政(zheng)策約束較(jiao)大(da)(da)(da)、集(ji)中出單占比高(gao)、季節(jie)(jie)性(xing)較(jiao)強(qiang)等特(te)點(dian)(dian)。在一些(xie)特(te)殊(shu)的(de)(de)應(ying)用場景，如財(cai)(cai)政(zheng)部(bu)下(xia)發《中央財(cai)(cai)政(zheng)農(nong)(nong)(nong)業(ye)保(bao)險(xian)(xian)(xian)保(bao)費補貼管理(li)辦法》要(yao)求(qiu)調整政(zheng)策性(xing)農(nong)(nong)(nong)險(xian)(xian)(xian)各(ge)險(xian)(xian)(xian)種的(de)(de)各(ge)級財(cai)(cai)政(zheng)補貼比例、農(nong)(nong)(nong)險(xian)(xian)(xian)集(ji)中承保(bao)錄(lu)(lu)單時進行海量(liang)影像上傳、農(nong)(nong)(nong)險(xian)(xian)(xian)理(li)賠大(da)(da)(da)客戶案(an)件抽檢要(yao)求(qiu)下(xia)載系(xi)(xi)統(tong)案(an)件清單及(ji)影像資(zi)(zi)料(liao)時，業(ye)務(wu)量(liang)巨大(da)(da)(da)且對時效性(xing)及(ji)數(shu)據(ju)質(zhi)量(liang)要(yao)求(qiu)都很高(gao)，分公司(si)往(wang)往(wang)需(xu)要(yao)抽調大(da)(da)(da)量(liang)業(ye)務(wu)人員集(ji)中在數(shu)天或(huo)(huo)數(shu)十天內(nei)加班加點(dian)(dian)完成(cheng)上萬件保(bao)單的(de)(de)錄(lu)(lu)單、批(pi)改(gai)(gai)或(huo)(huo)資(zi)(zi)料(liao)下(xia)載操作。為提高(gao)業(ye)務(wu)處理(li)效率，開發上線(xian)農(nong)(nong)(nong)險(xian)(xian)(xian)政(zheng)策性(xing)系(xi)(xi)數(shu)批(pi)改(gai)(gai)機(ji)器人、農(nong)(nong)(nong)險(xian)(xian)(xian)影像上傳機(ji)器人、農(nong)(nong)(nong)險(xian)(xian)(xian)理(li)賠資(zi)(zi)料(liao)下(xia)載機(ji)器人等農(nong)(nong)(nong)險(xian)(xian)(xian)業(ye)務(wu)機(ji)器人。根(gen)據(ju)業(ye)務(wu)需(xu)要(yao)，快速(su)部(bu)署數(shu)十名數(shu)字員工，完成(cheng)短期大(da)(da)(da)批(pi)量(liang)任務(wu)后(hou)再將機(ji)器人許可資(zi)(zi)源釋放，極大(da)(da)(da)地發揮了數(shu)字員工快速(su)部(bu)署、靈(ling)活(huo)適配(pei)業(ye)務(wu)的(de)(de)優勢，減輕了基(ji)層(ceng)一線(xian)的(de)(de)工作負擔。2022年7月，人保(bao)財(cai)(cai)險(xian)(xian)(xian)河南(nan)分公司(si)緊急開發上線(xian)并部(bu)署15個(ge)農(nong)(nong)(nong)險(xian)(xian)(xian)政(zheng)策性(xing)系(xi)(xi)數(shu)批(pi)改(gai)(gai)機(ji)器人，共運行2個(ge)月，完成(cheng)23890單批(pi)改(gai)(gai)工作，節(jie)(jie)約人力(li)498人天。

開創數字員工服務化模式的(de)運維機(ji)器人。人(ren)(ren)保(bao)財險(xian)湖南分(fen)公司(si)科技(ji)運(yun)營(ying)部率(lv)先(xian)提出“數(shu)字(zi)(zi)(zi)員(yuan)工(gong)(gong)服務(wu)(wu)(wu)化(hua)”理(li)念并(bing)(bing)實(shi)現落地，自(zi)主開發“小(xiao)(xiao)南同(tong)學(xue)”智(zhi)(zhi)能(neng)運(yun)維(wei)(wei)應用(yong)(yong)(yong)(yong)。一方(fang)面(mian)(mian)依托企業微信工(gong)(gong)作(zuo)臺作(zuo)為入(ru)口，接收用(yong)(yong)(yong)(yong)戶以(yi)自(zi)然(ran)語言方(fang)式下達的(de)(de)任務(wu)(wu)(wu)命令并(bing)(bing)轉化(hua)為數(shu)字(zi)(zi)(zi)員(yuan)工(gong)(gong)可以(yi)理(li)解(jie)(jie)的(de)(de)任務(wu)(wu)(wu)消息(xi)(xi)，另一方(fang)面(mian)(mian)將(jiang)各(ge)類運(yun)維(wei)(wei)數(shu)字(zi)(zi)(zi)員(yuan)工(gong)(gong)封裝成標準化(hua)服務(wu)(wu)(wu)以(yi)便調用(yong)(yong)(yong)(yong)。各(ge)數(shu)字(zi)(zi)(zi)員(yuan)工(gong)(gong)監聽消息(xi)(xi)，當監聽到歸(gui)屬于自(zi)身的(de)(de)任務(wu)(wu)(wu)消息(xi)(xi)時則自(zi)動執行，并(bing)(bing)將(jiang)執行結(jie)果返回“小(xiao)(xiao)南同(tong)學(xue)”前端(duan)并(bing)(bing)反饋用(yong)(yong)(yong)(yong)戶。“小(xiao)(xiao)南同(tong)學(xue)”將(jiang)日常運(yun)維(wei)(wei)中的(de)(de)重復工(gong)(gong)作(zuo)交給數(shu)字(zi)(zi)(zi)員(yuan)工(gong)(gong)自(zi)動解(jie)(jie)決，如車(che)險(xian)投(tou)保(bao)預(yu)確(que)認、重新轉保(bao)單(dan)、電子保(bao)單(dan)重發等，實(shi)現運(yun)維(wei)(wei)工(gong)(gong)作(zuo)的(de)(de)智(zhi)(zhi)能(neng)化(hua)自(zi)執行，并(bing)(bing)將(jiang)數(shu)字(zi)(zi)(zi)員(yuan)工(gong)(gong)的(de)(de)調用(yong)(yong)(yong)(yong)無縫融入(ru)運(yun)維(wei)(wei)工(gong)(gong)作(zuo)場景(jing)，支(zhi)持數(shu)字(zi)(zi)(zi)員(yuan)工(gong)(gong)的(de)(de)自(zi)定義擴展和快速接入(ru)，實(shi)現“數(shu)字(zi)(zi)(zi)員(yuan)工(gong)(gong)服務(wu)(wu)(wu)化(hua)”模式。人(ren)(ren)保(bao)財險(xian)科技(ji)運(yun)營(ying)部基(ji)于湖南分(fen)公司(si)創新探索，打造統頒“人(ren)(ren)保(bao)小(xiao)(xiao)智(zhi)(zhi)”智(zhi)(zhi)能(neng)運(yun)維(wei)(wei)應用(yong)(yong)(yong)(yong)并(bing)(bing)進行全國推(tui)廣(guang)。

數字(zi)員工(gong)具有廣闊的應(ying)用前景。下一(yi)(yi)步(bu)，人保財險科技運營(ying)部將持續深化(hua)數字(zi)員工(gong)建(jian)設和應(ying)用，協(xie)同(tong)業務部門(men)及分公(gong)司進(jin)(jin)一(yi)(yi)步(bu)挖掘、擴大應(ying)用場(chang)景，優化(hua)、完(wan)善數字(zi)員工(gong)運營(ying)管理平臺支撐能(neng)力，提升數字(zi)員工(gong)效(xiao)能(neng)以及運行穩定性，加強總分協(xie)同(tong)、技術(shu)培(pei)訓和流程推(tui)廣，深化(hua)RPA與(yu)AI等技術(shu)融合創(chuang)新，積極與(yu)同(tong)業交流學習先進(jin)(jin)經(jing)驗，探索(suo)應(ying)用RPM（機器人流程挖掘）技術(shu)發掘更多(duo)應(ying)用場(chang)景，進(jin)(jin)一(yi)(yi)步(bu)促進(jin)(jin)降本增效(xiao)，助(zhu)力運營(ying)向(xiang)自動化(hua)和智能(neng)化(hua)轉型。