Critical success factors in BPM implementation: Creating AI supported decision engine for the business

Mas e Deg ee P og am in In o ma ion Sys ems and Technologies Managemen C i ical success ac o s in BPM implemen a ion: C ea ing AI suppo ed decision engine o he business. Vic o Lipo Mas e Thesis p esen ed as pa ial equi emen o ob aining he Mas e Deg ee in In o ma ion Managemen NOVA In o ma ion Managemen School Ins i u o Supe io de Es a ís ica e Ges ão de In o mação Uni e sidade No a de Lisboa MGI NOVA In o ma ion Managemen School Ins i u o Supe io de Es a ís ica e Ges ão de In o mação Uni e sidade No a de Lisboa C i ical success ac o s in BPM implemen a ion: C ea ing AI suppo ed decision engine o he business. by Vic o Lipo Mas e Thesis p esen ed as he pa ial equi emen o ob aining a Mas e 's deg ee in In o ma ion Managemen , specializa ion in In o ma ion Sys ems and Technologies Managemen Supe ised by Ped o Manuel Maia Mal a, PhD NOVA In o ma ion Managemen School July, 2024 i STATEMENT OF INTEGRITY I he eby decla e ha ing conduc ed his academic wo k wi h in eg i y. I con i m ha I ha e no used plagia ism o any o m o undue use o in o ma ion o alsi ica ion o esul s along he p ocess leading o i s elabo a ion. I u he decla e ha I ha e ully acknowledged he Rules o Conduc and Code o Hono om he NOVA In o ma ion Managemen School. [Lisboa, 13/07/2024] ii ABSTRACT In oday's apidly e ol ing business en i onmen , achie ing ope a ional e iciency, anspa ency, and ma ke esponsi eness is c ucial o success. Business P ocess Managemen (BPM) is a c i ical ool in his pu sui , encompassing he design, con igu a ion, enac men , and analysis o business p ocesses o d i e con inuous imp o emen . Recen ad ancemen s in Gene a i e P e- ained T ans o me (GPT) models o e new oppo uni ies o enhance BPM h ough imp o ed decision-making and p og ess- acking capabili ies. Howe e , he e is a signi ican gap in he li e a u e conce ning he sys ema ic de elopmen o AI-powe ed decision engines o suppo BPM implemen a ion. This hesis add esses his gap by pa ially employing Design Science Resea ch Me hodology (DSRM) wi h some limi a ions desc ibed in he documen o de elop a decision engine a i ac ailo ed o BPM adop ion. The esea ch u ilizes concep o C i ical Success Fac o s (CSFs) essen ial o BPM, inco po a ing AI o au oma e and op imize decision-making p ocesses. The me hodology in ol es a ho ough li e a u e e iew, he concep ualiza ion and design o a decision engine model, and he de elopmen o a p o o ype. This p o o ype is e alua ed h ough expe e iews, p o iding c i ical insigh s in o i s e ec i eness and p ac ical applicabili y. The indings indica e ha he in eg a ion o AI, pa icula ly GPT models, can signi ican ly enhance BPM by p o iding dynamic insigh s and au oma ing complex decision-making p ocesses. The de eloped decision engine p o o ype o e s a s uc u ed app oach o BPM adop ion, ailo ed o di e en o ganiza ional s ages and suppo ed by a comp ehensi e lis o CSFs. This esea ch con ibu es o bo h academic knowledge and p ac ical applica ions, o e ing a ounda ional amewo k o o ganiza ions o enhance hei BPM e o s and ad ance he dialogue on da a-d i en decision-making in business p ocess implemen a ion. KEYWORDS Business P ocess Managemen ; A i icial In elligence; Decision Engine; C i ical Success Fac o s Sus ainable De elopmen Goals (SDG): iii TABLE OF CONTENTS S a emen o In eg i y ...................................................................................................... i Abs ac ............................................................................................................................ ii Lis o Figu es .................................................................................................................... Lis o Tables .................................................................................................................... i Lis o Abb e ia ions and Ac onyms ............................................................................... ii 1. In oduc ion ................................................................................................................ 1 2. Li e a u e e iew ........................................................................................................ 2 2.1. Key Concep s and De ini ions o AI- ela ed hings ............................................. 2 2.1.1. A i icial In elligence (AI) .............................................................................. 2 2.1.2. La ge Language Models (LLMs) ..................................................................... 2 2.1.3. GPT (Gene a i e P e- ained T ans o me ) .................................................. 3 2.1.4. Business P ocess Managemen (BPM) .......................................................... 3 2.1.5. C i ical Success Fac o s (CSFs) ...................................................................... 3 2.1.6. Decision Engine ............................................................................................. 3 2.1.7. Sen imen Analysis ........................................................................................ 3 2.2. De ailed Explana ions o AI Componen s ............................................................ 3 2.2.1. Machine Lea ning (ML) ................................................................................. 3 2.2.2. Na u al Language P ocessing (NLP) .............................................................. 4 2.2.3. Deep Lea ning ............................................................................................... 4 2.3. Iden i ica ion o cu en s a e o bpm adop ion o legal en i y ........................... 4 2.4. Role o AI in BPM ................................................................................................. 6 3. Me hodology .............................................................................................................. 9 3.1. P oblem Iden i ica ion ......................................................................................... 9 3.2. Objec i es o a Solu ion ....................................................................................... 9 3.3. Design and De elopmen ..................................................................................... 9 3.4. Summa y o An icipa ed Ou comes ................................................................... 10 4. Empi ical S udy ......................................................................................................... 12 4.1. App oach o decision engine p o o ype ............................................................ 12 4.2. Decision engine p o o ype ................................................................................. 13 4.3. Simula ed Example o p ac ical applica ion. ...................................................... 16 5. Resul s and discussion .............................................................................................. 19 5.1. Exposi ion o Empi ical Resul s .......................................................................... 19 i 5.2. Valida ion o Empi ical Resul s .......................................................................... 19 5.2.1. CEO ............................................................................................................. 20 5.2.2. CFO .............................................................................................................. 20 5.2.3. Manage ...................................................................................................... 21 5.2.4. Summa y o he eedback: .......................................................................... 21 5.3. Discussion .......................................................................................................... 22 5.4. DE P o o ype Implemen a ion Adjus men s Based on Expe Sugges ions ....... 22 6. Conclusions and u u e wo ks ................................................................................... 24 6.1. Syn hesis o he Wo k Done wi h Highligh o he Recommenda ions Made ... 24 6.1.1. Key Recommenda ions: .............................................................................. 24 6.2. Limi a ions o he Wo k Done ............................................................................ 25 6.3. Fu he Wo k Possible o Do .............................................................................. 25 Bibliog aphical Re e enczoes ......................................................................................... 26 Appendix A - E hics Commi ee Repo ......................................................................... 28 Appendix B – Decision engine p o o ype ...................................................................... 29 Appendix C - Expe Ques ionnai e ............................................................................... 30 LIST OF FIGURES Figu e 2.4.1 - BPM li ecycle en iched wi h AI ............................................................................ 7 Figu e 4.1.1 - DE p o o ype logic ............................................................................................. 12 Figu e 4.2.1 - BPM Adop ion S age P og ession ...................................................................... 15 Figu e 4.2.2 - CSF Sco ing App oach ........................................................................................ 15 i LIST OF TABLES Table 2.3.1 - Top 10 ac o s a ec ing BMP implemen a ion (Cas o e al., 2019) .................... 5 Table 2.3.2 - CSFs o di e en s ages o BPM adop ion (Buh e al., 2015) ............................... 6 Table 4.2.1 - BPM Adop ion S ages ......................................................................................... 13 Table 4.2.2 - C i ical Success Fac o s (CSFs) o S age 1 .......................................................... 14 Table 4.3.1 - BPM adop ion simula ion ................................................................................... 16 Table 5.4.1 - DE adjus men s based on eedback .................................................................... 22 ii LIST OF ABBREVIATIONS AND ACRONYMS AI A i icial In elligence BPM Business P ocess Managemen CSF C i ical Success Fac o s DSRM Design Science Resea ch Me hodology GPT Gene a i e P e- ained T ans o me LLM La ge Language Models NLP Na u al Language P ocessing ROI Re u n on In es men SDG Sus ainable De elopmen Goals DE Decision Engine 7 domains and asks (Teubne e al., 2023) and GPT-like models, ha e signi ican ly enhanced he abili y o analyze and iden i y bo h explici and implici p ocesses wi hin an o ganiza ion. I means ha you can apply GPT-like models o s uc u ed and uns uc u ed da a o dig he p ocesses o de elop solu ion o he p oblem. In he esea ch done by (Weinzie l e al., 2024), he e a e h ee equen examples o using ML: • p o iding decision suppo h ough p edic ions • disco e ing accu a e p ocess models • imp o ing esou ce alloca ion. Keeping his in mind, esea che s ha e de eloped se e al BPM-speci ic solu ions ha ex ac in o ma ion om ex ual documen s using Na u al Language P ocessing echniques. (G ohs e al., 2023) In his esea ch, G ohs e al., 2023 illus a ed how LLM can be u ilized o p ac ical BPM asks which equi ed ex ual documen s as inpu . Cha GPT4 was u ilized o his ask. Fo all h ee asks, GPT4 pe o ms simila ly o o be e han he benchma k, i.e., speci ic applica ions o he espec i e ask (G ohs e al., 2023). The e a e also inc easing numbe o se ices o Con e be ween P ocess Models and Na u al Language Tex u ilizing g aphical modeling languages. I allows s akeholde s o collec i ely desc ibe execu able business p ocesses on a concep ual le el. (F ey ag e al., n.d.). Comp ehensi e li e a u e e iew being done on he opic o “Machine lea ning in business p ocess managemen ” by (Weinzie l e al., 2024). Resul ed in en iched BPM li ecycle BPM li ecycle o Dumas e al. (2018). Which gi es highe le el p ac ical iew pe spec i e on he a eas, whe e AI can be applied. Figu e 2.4.1 - BPM li ecycle en iched wi h AI 8 F om he pe spec i e o ou esea ch, i helps o unde s and be e he ongoing p ocesses in he o ganiza ion and ack he p ocesses which a e adop ed o BPM. Machine Lea ning (ML) has been pa icula ly use ul in imp o ing ce ain pa s o Business P ocess Managemen (BPM). These imp o emen s mos ly happen in a eas whe e he e is a lo o da a, such as disco e ing new p ocesses, analyzing hem, and keeping an eye on hem o e ime. In hese a eas, da a abou business ac i i ies is usually a ailable in de ailed eco ds, which ML can easily use (Sa ke , 2021; Weinzie l e al., 2024). Howe e , esea che s sugges ha we could make e en mo e p og ess by using echniques om ields like na u al language p ocessing (NLP) and compu e ision. NLP helps compu e s unde s and and wo k wi h human language, while compu e ision helps hem unde s and images and ideos. By combining hese echniques wi h ML, we could c ea e new and be e ways o manage business p ocesses ha go beyond jus using da a logs. This could lead o inno a i e solu ions ha a e mo e e ec i e and e sa ile. Weinzie l e al., 2024 also p opose o conside he en e p ise p ocess ne wo k ins ead o isola ed business p ocesses. In doing ha , ML applica ions ecei e inpu da a om mul iple da a sou ces, including con ol- low in o ma ion om di e en business p ocesses and p ocess con ex in o ma ion ela ed o he business p ocesses. Le e aging LLMs and GPT models acili a es comp ehensi e p ocess iden i ica ion, dynamic oadmap de elopmen , eal- ime implemen a ion moni o ing, and he acking o quali a i e me ics. These capabili ies add ess signi ican gaps in adi ional BPM app oaches, pa ing he way o mo e e ec i e and e icien sys ems. By con inuously in eg a ing AI-d i en insigh s, o ganiza ions can ensu e hei BPM ini ia i es a e no only success ul bu also adap able o u u e challenges and oppo uni ies (G ohs e al., 2023; Weinzie l e al., 2024). In addi ion o adi ional pe o mance me ics, acking quali a i e me ics such as sen imen and in ol emen (how people eel and how in ol ed hey a e in hei wo k) is i al o comp ehensi e BPM e alua ion. AI echnologies, including sen imen analysis and engagemen acking ools, can analyze quali a i e da a o p o ide deepe insigh s in o employee mo ale, cus ome sa is ac ion, and o e all o ganiza ional engagemen . These insigh s p o ide a deepe unde s anding o he wo kplace a mosphe e and cus ome expe iences, allowing o ganiza ions o pinpoin a eas ha need imp o emen and ake speci ic ac ions o enhance o e all BPM esul s. 9 3. METHODOLOGY This sec ion ou lines he me hodology adop ed o his esea ch, aiming o iden i y and u ilize C i ical Success Fac o s (CSFs) wi hin BPM o de elop a decision engine a i ac ha suppo s BPM implemen a ion. The ini ial idea was o apply Design Science Resea ch Me hodology (DSRM) as a amewo k o his s udy (Hasan & Sha i zadeh, 2020; Hunzike & Blankenagel, 2024). Howe e , due o he lack o access o a eal-wo ld company, i was no easible o accomplish s ages o DSRM such as Demons a ion and E alua ion ully. Ne e heless, some pa s o DSRM we e applied, such as P oblem Iden i ica ion, Objec i es o a Solu ion, Design and De elopmen . De ailed desc ip ion p o ided below: 3.1. PROBLEM IDENTIFICATION Objec i e: The p ima y goal a his s age is o iden i y he cu en challenges and gaps in BPM implemen a ion. This in ol es a deep di e in o exis ing li e a u e and esea ch o pinpoin a eas whe e BPM p ac ices a e lacking, pa icula ly ocusing on he sys ema ic de elopmen o AI-suppo ed decision engines. The objec i e is o es ablish a clea unde s anding o he p oblems aced by o ganiza ions in adop ing and op imizing BPM sys ems. App oach: A comp ehensi e li e a u e e iew was conduc ed o ga he insigh s in o he cu en s a e o BPM adop ion, common challenges, and he po en ial ole o AI in enhancing BPM. This in ol ed analyzing a ious schola ly a icles, indus y epo s, and case s udies o iden i y ecu ing hemes and issues. The li e a u e e iew also included an examina ion o exis ing AI echnologies and hei applica ions in BPM, p o iding a ounda ional unde s anding o how AI can be le e aged o add ess iden i ied challenges. A his s age we iden i ied impo ance o CSFs o BPM adop ion as well as b eakdown o BPM adop ion o di e en S ages (s a es o o ganiza ion). 3.2. OBJECTIVES OF A SOLUTION Objec i e: The nex s ep is o de ine he goals and equi emen s o he AI-suppo ed decision engine. This in ol es se ing clea , measu able objec i es ha he decision engine aims o achie e. The objec i e is o ensu e ha he p oposed solu ion add esses he iden i ied challenges and gaps in BPM implemen a ion e ec i ely. App oach: Insigh s om he li e a u e e iew and expe in e iews we e u ilized o ou line he objec i es. The objec i es we e e ined h ough i e a i e eedback, ensu ing hey a e aligned wi h he needs o o ganiza ions and he capabili ies o AI echnologies. This app oach ensu ed ha he solu ion is bo h heo e ically sound and p ac ically ele an . 3.3. DESIGN AND DEVELOPMENT Objec i e: The p ima y goal he e is o de elop he decision engine a i ac based on he iden i ied CSFs and he de ined objec i es. This in ol es a de ailed design and 10 de elopmen p ocess, ensu ing ha he decision engine is obus , scalable, and capable o add essing he iden i ied challenges in BPM implemen a ion. S age 1: Iden i ica ion o Cu en S a e o BPM Adop ion App oach: The cu en BPM adop ion s age o he o ganiza ion was assessed using a s uc u ed amewo k. This amewo k was de i ed om he esea ch o Buh e al. (2015), which ca ego izes BPM adop ion in o a ious s ages and iden i ies associa ed CSFs o each s age. S age 2: Iden i y CSFs o Respec i e S a e App oach: The lis o c i ical success ac o s (CSFs) was condensed om Buh e al. (2015) esea ch o ocus on he mos ele an ac o s o each BPM adop ion s age. This in ol ed p io i izing CSFs ha ha e he highes impac on BPM success and a e mos applicable o he o ganiza ion's cu en s a e. This p ocess included: • Re iewing he de ailed desc ip ions and jus i ica ions o each CSF. • E alua ing hei applicabili y based on he o ganiza ion's speci ic con ex and needs. • Engaging wi h subjec ma e expe s o alida e he selec ion and p io i iza ion o CSFs. S age 3: De elop a Sco ing App oach o Respec i e S a e App oach: The sco ing app oach was de eloped wi h he suppo o GPT-4 h ough a se ies o p omp s and p ac ical expe ience. This p ocess in ol ed: • De ining clea c i e ia o each CSF based on li e a u e and expe inpu . • De eloping a a ing scale (e.g., 1 o 5) o each c i e ion o quan i y he le el o achie emen o pe o mance. • Tes ing he sco ing app oach h ough simula ed scena ios and expe alida ion o ensu e i s eliabili y and alidi y. P omp s example: • Please sugges lis o c i e ia which can be used o sco ing {p oposed sco ing app oach}. • Wha kind o scales can be applied o {measu emen c i e ia}. Responses we e c i ically analyzed and adop ed o he opic. S age 4: C i ical E alua ion App oach: In e iews wi h expe s ac oss di e en le els o he co po a e hie a chy we e conduc ed o e alua e he p o o ype om a ious pe spec i es. This included: • Conduc ing s uc u ed in e iews wi h CEO le el, CFO le el and line manage s o ga he di e se insigh s. • Analyzing he eedback o iden i y common hemes, s eng hs, and a eas o imp o emen in he decision engine. 3.4. SUMMARY OF ANTICIPATED OUTCOMES The an icipa ed ou comes o his esea ch include: 11 • A lis o C i ical Success Fac o s (CSFs) o BPM implemen a ion, ailo ed o di e en s ages o BPM adop ion. • A heo e ical model o an AI-suppo ed decision engine ha inco po a es hese CSFs, p o iding a s uc u ed app oach o enhance BPM p ac ices. • C i ical insigh s om expe s ac oss di e en le els o expe ise in BPM. This esea ch aspi es o con ibu e o bo h academic knowledge and p ac ical applica ions in BPM. By p o iding a ounda ional amewo k o AI-enhanced BPM implemen a ion, i aims o suppo o ganiza ions in achie ing highe le els o ope a ional e iciency, anspa ency, and ma ke esponsi eness. 12 4. EMPIRICAL STUDY 4.1. APPROACH TO DECISION ENGINE PROTOTYPE De elopmen o Decision Engine (DE) can add ess complexi y and add anspa ency o he p ocess o BPM adop ion. As pe de ini ion, decision engines a e sophis ica ed so wa e sys ems designed o au oma e he decision-making p ocesses wi hin an o ganiza ion. I ope a es by analyzing inpu da a h ough a se o p ede ined ules, algo i hms, o machine lea ning models o make decisions o ecommenda ions wi hou human in e en ion (pu ely based on a ailable da a). These engines a e pa icula ly aluable in scena ios whe e as , consis en , and accu a e decisions a e c ucial, such as in c edi sco ing, aud de ec ion, cus ome se ice, and many o he ope a ional p ocesses. Fo his pape , we decided s ep aside om gene alized ac o s and align decision engine wi h di e en s ages o BPM adop ion. To do his, we de eloped he p o o ype in he ollowing logic: Figu e 4.1.1 - DE p o o ype logic The idea behind is ha be o e we s a , we should unde s and whe e we a e now. When we unde s and whe e we a e, we can decide wha o do nex and how o do i . 13 4.2. DECISION ENGINE PROTOTYPE F om ou pe spec i e decision engine p o o ype should be adop able and p ac ical. I should guide he end use h ough he di e en s ages; he e o e, i should help answe ing wo ini ial ques ions: • Whe e he o ganiza ion is now? • Wha should i do nex ? To answe o he i s ques ion he p o o ype includes desc ip ion o BPM adop ion s ages as pe Buh e al., 2015. Iden i ica ion, and Mo ing o wa d ac ions we e de i ed wi h suppo o Cha GPT4 using p omp s like: • based on you expe ience, how can o ganiza ion be iden i ied a {S age} • please sugges possible ac ions o he o ganiza ion o ad ance u he and summa ized o he able below: Table 4.2.1 - BPM Adop ion S ages S age Desc ip ion Iden i ica ion Mo ing Fo wa d Ac ions Awa eness and Unde s anding o BPM O ganiza ions ecognize he impo ance o BPM. P ocess imp o emen is ad hoc, eac i e, and uns uc u ed. BPM is a new concep o he e is limi ed unde s anding o p ocess managemen p inciples o ganiza ion-wide. In oduce o mal BPM aining and es ablish a sense o u gency o p ocess imp o emen . Desi e o Adop BPM O ganiza ions a e mo i a ed o adop BPM and a e beginning o plan hei ini ia i es. Managemen shows in e es in BPM, and ini ial esou ces a e alloca ed. Secu e op managemen suppo , appoin a p ojec champion, and de ine business d i e s. BPM P ojec s O ganiza ions execu e speci ic BPM p ojec s wi h de ined objec i es and plans. Clea objec i es, plans, and ex e nal consul an guidance a e in place. Communica e objec i es, seek p o essional guidance, and mo i a e eam membe s. BPM P og am BPM is in eg a ed in o he o ganiza ion’s b oade s a egy and p og ams. Con inuous op managemen suppo , p o essional guidance, and e ec i e communica ion a e e iden . Ensu e ongoing suppo , main ain consul an pa ne ships, and iden i y KPIs. P oduc iza ion o BPM BPM p ac ices a e s anda dized and Key pe o mance indica o s (KPIs) a e Focus on aining, secu ing op 14 p oduc ized wi hin he o ganiza ion. de ined, employees a e educa ed and mo i a ed, and BPM is embedded in decision-making. managemen endo semen , and e ining p ocesses based on KPIs. Fo example: S age 1: Awa eness and Unde s anding o BPM • Desc ip ion: O ganiza ions a his s age ha e begun o ecognize he impo ance o BPM. P ocess imp o emen is ypically ad hoc, eac i e, and uns uc u ed. The o ganiza ion ecognizes BPM as a po en ial alue-add and begins o unde s and i s p inciples and bene i s. • Iden i ica ion: I BPM is a new concep o i he e is limi ed unde s anding o p ocess managemen p inciples o ganiza ion-wide, i 's likely a his s age. • Mo ing Fo wa d: In oduce o mal BPM aining and es ablish a sense o u gency o p ocess imp o emen . I he desc ip ion ma ches he ype o o ganiza ion, he o ganiza ion can become awa e o ocus poin s and sugges ions ele an o he s age. In ou example: • Focus: Educa ion on BPM impo ance and i s impac on pe o mance. • Educa ion and Communica ion: Conduc wo kshops and aining sessions o educa e s akeholde s abou he bene i s and p inciples o BPM. • Visible Execu i e Suppo : Secu e and showcase execu i e endo semen o highligh he impo ance o BPM wi hin he o ganiza ion. • Case S udies and Success S o ies: Sha e case s udies o success ul BPM implemen a ions o illus a e po en ial gains and encou age buy-in. The same applies o po en ial p oblems. As a esul , he o ganiza ion ge s a lis o CSFs o conside . In ou example, o he i s s age o BPM adop ion, CSFs include empowe men o employees, cus ome ocus, and openness o changes (Buh e al., 2015). Table 4.2.2 - C i ical Success Fac o s (CSFs) o S age 1 CSF Focus Poin s Ac ions Empowe men o Employees T aining and wo kshops o enhance unde s anding o BPM. Conduc mon hly BPM aining sessions and in e ac i e wo kshops. 15 Cus ome Focus Implemen eedback loops o align p ocesses wi h cus ome needs. In oduce su eys and eedback o ms pos -cus ome in e ac ions. Openness o Changes Encou age a cul u e o adap abili y and con inuous imp o emen . De elop and communica e a s uc u ed change managemen plan, implemen a ecogni ion p og am o success ul changes. By knowing CSFs, he o ganiza ion can de elop sco ing app oaches, measu emen c i e ia, scales, e c. Below is a lowcha ep esen ing he p og ession h ough di e en s ages o BPM adop ion, highligh ing key ac ions and ocus poin s a each s age. Figu e 4.2.1 - BPM Adop ion S age P og ession A ada cha showing he sco ing o di e en CSFs helps isualize he s eng hs and a eas o imp o emen o each ac o . Figu e 4.2.2 - CSF Sco ing App oach Including hese ables and cha s enhances he cla i y and p ac icali y o he p o o ype, making i easie o o ganiza ions o na iga e h ough he s ages o BPM adop ion. 16 4.3. SIMULATED EXAMPLE OF PRACTICAL APPLICATION. Wi h he suppo o Cha GPT4 we ha e done se ies o p ac ical simula ions on how i can be applied o he eal-wo ld scena ios. Company P o ile: Name: Tech Inno a o s Inc. Indus y: Technology Solu ions Cu en BPM Adop ion S age: Awa eness and Unde s anding o BPM The company decides o s a BPM adop ion smoo hly by ocusing on wo key p ocesses: he Cus ome Suppo P ocess and he P oduc De elopmen P ocess. P omp s used: • Please gene a e lis o possible asks o {s ep} o {ac ion / ask} o {s age} • Wha possible esul s/ ou comes can be o {s ep} o {ac ion / ask} o {s age} Table 4.3.1 - BPM adop ion simula ion S ep Ac ion/Task Expec ed Resul 1. Ini ial Assessmen Su eys Conduc in e nal su eys o ga he insigh s om employees. Clea unde s anding o cu en BPM knowledge and engagemen . In e iews Conduc in e iews wi h key s akeholde s. Iden i y pain poin s and oppo uni ies o imp o emen . Documen Re iew Analyze exis ing p ocess documen a ion, aining eco ds, and pe o mance me ics. Iden i y gaps and a eas needing enhancemen . Ou come The assessmen e ealed ha while he e was an acknowledgmen o he need o BPM, he p ocess imp o emen ini ia i es we e ad hoc, eac i e, and uns uc u ed. This placed Tech Inno a o s in he "Awa eness and Unde s anding o BPM" s age. 2. De ine CSFs Empowe men o Employees Enhance employees' unde s anding o BPM and encou age con ibu ions. Imp o ed employee engagemen and pa icipa ion in BPM ini ia i es. Cus ome Focus Implemen mechanisms o ga he and ac on cus ome eedback. Be e alignmen o cus ome needs wi h suppo p ocesses. 23 cus ome eedback, and os e adap abili y. and inc eased lexibili y. clea inancial me ics. inancial cla i y, and adap abili y. De elop Sco ing App oach De elop c i e ia and a ing scales o each CSF based on li e a u e and expe inpu . S uc u ed e alua ion o BPM e ec i eness. Manage : Ensu e sco ing is simple and ac ionable. P ac ical and easy- o-apply sco ing sys em o e alua ing BPM e ec i eness. Implemen Decision Engine Conduc BPM aining sessions, in oduce eedback mechanisms, and communica e change managemen plans. Inc eased unde s anding o BPM, enhanced cus ome eedback in eg a ion, and highe adap abili y o change. CEO: Balance AI insigh s wi h human in e ac ion; CFO: Ensu e cos - bene i analysis; Manage : Suppo ca ee de elopmen . Imp o ed BPM unde s anding, be e eedback in eg a ion, cos - e ec i e decisions, and enhanced employee g ow h. E alua ion and Resul s Measu e imp o emen s in employee engagemen , cus ome sa is ac ion, and adap abili y. Demons a ed imp o emen s in BPM p ac ices and s akeholde engagemen . CFO: Include inancial KPIs; Manage : Regula eedback sessions. Comp ehensi e e alua ion showing enhanced BPM p ac ices, inancial pe o mance, and s akeholde engagemen . This esea ch con ibu es o bo h academic discou se and p ac ical applica ions in BPM. By de eloping and alida ing an AI-suppo ed decision engine p o o ype, i o e s a ounda ional amewo k o o ganiza ions o enhance hei BPM e o s. The amewo k can be used in any ype o o ganiza ion and on di e en s ages o de elopmen , bu p ac ical applica ion should be suppo ed by u he esea ches. The in eg a ion o AI echnologies and human-cen ic design p inciples ensu es ha BPM ini ia i es a e e ec i e, adap able, and aligned wi h o ganiza ional goals. Fu u e esea ch could ocus on e ining he decision engine based on b oade indus y eedback and explo ing i s applica ion in di e en o ganiza ional se ings. 24 6. CONCLUSIONS AND FUTURE WORKS In his inal chap e , we syn hesize he wo k done in his esea ch, highligh he ecommenda ions made, discuss he limi a ions encoun e ed, and sugges possible di ec ions o u he esea ch. 6.1. SYNTHESIS OF THE WORK DONE WITH HIGHLIGHT TO THE RECOMMENDATIONS MADE This hesis explo ed he in eg a ion o AI-suppo ed decision engines in o Business P ocess Managemen (BPM) o enhance BPM adop ion and implemen a ion. The p ima y con ibu ions o his esea ch a e: Li e a u e Re iew: A comp ehensi e li e a u e e iew iden i ied C i ical Success Fac o s (CSFs) c ucial o BPM implemen a ion and examined he ole o AI, pa icula ly La ge Language Models (LLMs) like GPT, in BPM. De elopmen o a Decision Engine P o o ype: Using some o he componen s o Design Science Resea ch Me hodology, we de eloped a decision engine p o o ype ailo ed o di e en s ages o BPM adop ion. The engine was designed o guide o ganiza ions h ough awa eness, adop ion, p ojec execu ion, and p oduc iza ion s ages, p o iding s age- speci ic ecommenda ions. Empi ical S udy: The p o o ype was es ed h ough a simula ed case s udy o Tech Inno a o s Inc., demons a ing i s p ac ical applica ion and e ec i eness in imp o ing BPM p ac ices. Expe Valida ion: In e iews wi h indus y expe s alida ed he app oach and p o ided c i ical eedback o imp o emen . The ecommenda ions om expe s included managing expec a ions, inco po a ing inancial me ics, ensu ing clea communica ion, and balancing AI-d i en insigh s wi h human judgmen . 6.1.1. Key Recommenda ions: Realis ic Communica ion: Clea ly communica e he bene i s and limi a ions o BPM o a oid o e hyping and o build a solid ounda ion. Leade ship Commi men : Ensu e s ong and empa he ic leade ship o d i e BPM ini ia i es. Financial Me ics: In eg a e inancial pe o mance me ics o demons a e ROI and secu e execu i e suppo . Human In e ac ion: Balance AI insigh s wi h human-cen ic ea u es o build us and acili a e engagemen . 25 T aining and Educa ion: Conduc p ac ical aining and wo kshops o enhance employee unde s anding. 6.2. LIMITATIONS OF THE WORK DONE Se e al limi a ions we e encoun e ed du ing his esea ch: Lack o Real-Wo ld Company Access: The inabili y o access a eal-wo ld company limi ed he scope o empi ical alida ion. The simula ed case s udy p o ided aluable insigh s bu lacked he complexi y o eal-wo ld implemen a ion. Scope o Expe In e iews: While expe eedback was c ucial, he sample size was small. A b oade ange o indus y pe spec i es could p o ide mo e comp ehensi e alida ion. P o o ype Tes ing: The decision engine p o o ype was es ed in a con olled en i onmen . Real-wo ld es ing migh e eal addi ional challenges and a eas o e inemen . 6.3. FURTHER WORK POSSIBLE TO DO Fu u e esea ch could add ess he limi a ions and expand on he indings o his hesis: Real-Wo ld Implemen a ion: Conduc case s udies wi h eal-wo ld companies o alida e and e ine he decision engine p o o ype in di e se o ganiza ional se ings. B oade Expe Feedback: Engage wi h a la ge and mo e di e se g oup o indus y expe s o ob ain comp ehensi e alida ion and insigh s. Ad anced AI In eg a ion: Explo e he in eg a ion o mo e ad anced AI echniques, such as ein o cemen lea ning and hyb id AI models, o enhance he decision engine's capabili ies. Longi udinal S udies: Conduc longi udinal s udies o assess he long- e m impac o AI- suppo ed decision engines on BPM adop ion and pe o mance. Cus omiza ion o Di e en Indus ies: De elop indus y-speci ic e sions o he decision engine o add ess unique challenges and equi emen s in sec o s such as heal hca e, inance, and manu ac u ing. By add essing he limi a ions and pu suing u he esea ch, we can con inue o ad ance he ield and suppo o ganiza ions in achie ing g ea e ope a ional e iciency and ma ke esponsi eness. 26 BIBLIOGRAPHICAL REFERENCZOES Buh, B., Ko ačič, A., & Indiha Š embe ge , M. (2015). C i ical success ac o s o di e en s ages o business p ocess managemen adop ion – a case s udy. Economic Resea ch- Ekonomska Is aži anja, 28(1), 243–258. h ps://doi.o g/10.1080/1331677X.2015.1041776 Cas o, B. K. D. A., D esch, A., & Vei , D. R. (2019). Key c i ical success ac o s o BPM implemen a ion: A heo e ical and p ac ical iew. Business P ocess Managemen Jou nal, 26(1), 239–256. h ps://doi.o g/10.1108/BPMJ-09-2018-0272 Dumas, M., La Rosa, M., Mendling, J., & Reije s, H. A. (2018). Fundamen als o Business P ocess Managemen . Sp inge Be lin Heidelbe g. h ps://doi.o g/10.1007/978-3-662-56509- 4 F ey ag, T., Kanzle , B., Lege , N., & Semling, D. (n.d.). NLP as a Se ice: An API o Con e be ween P ocess Models and Na u al Language Tex . G ohs, M., Abb, L., Elsayed, N., & Rehse, J.-R. (2023). La ge Language Models can accomplish Business P ocess Managemen Tasks (a Xi :2307.09923). a Xi . h p://a xi .o g/abs/2307.09923 Han, J., Pei, J., & Tong, H. (2022). Da a mining: Concep s and echniques. Mo gan kau mann. Hasan, R., & Sha i zadeh, R. (2020). Design Science Resea ch Me hodology As a Solu ion- O ien ed Me hodology. Hunzike , S., & Blankenagel, M. (2024). Design Science Resea ch Design (pp. 97–116). h ps://doi.o g/10.1007/978-3-658-42739-9_6 Khu ana, D., Koli, A., Kha e , K., & Singh, S. (2023). Na u al language p ocessing: S a e o he a , cu en ends and challenges. Mul imedia Tools and Applica ions, 82(3), 3713– 3744. h ps://doi.o g/10.1007/s11042-022-13428-4 Liu, B. (2020). In oduc ion. In Sen imen Analysis: Mining Opinions, Sen imen s, and Emo ions (pp. 1–17). Camb idge Uni e si y P ess. Sa ke , I. H. (2021). Machine Lea ning: Algo i hms, Real-Wo ld Applica ions and Resea ch Di ec ions. SN Compu e Science, 2(3), 160. h ps://doi.o g/10.1007/s42979-021- 00592-x Teubne , T., Fla h, C. M., Weinha d , C., Van De Aals , W., & Hinz, O. (2023). Welcome o he E a o Cha GPT e al.: The P ospec s o La ge Language Models. Business & In o ma ion Sys ems Enginee ing, 65(2), 95–101. h ps://doi.o g/10.1007/s12599-023-00795-x an de Aa, H., Ca mona, J., Leopold, H., Mendling, J., & Pad ó, L. (2018, Augus ). Challenges and Oppo uni ies o Applying Na u al Language P ocessing in Business P ocess Managemen . Van De Aals , W. M. P. (2013). Business P ocess Managemen : A Comp ehensi e Su ey. ISRN So wa e Enginee ing, 2013, 1–37. h ps://doi.o g/10.1155/2013/507984 Weinzie l, S., Zilke , S., Dunze , S., & Ma zne , M. (2024). Machine lea ning in business p ocess managemen : A sys ema ic li e a u e e iew. Expe Sys ems wi h Applica ions, 253, 124181. h ps://doi.o g/10.1016/j.eswa.2024.124181 27 Weske, M. (2012). In oduc ion. In M. Weske (Ed.), Business P ocess Managemen : Concep s, Languages, A chi ec u es (pp. 3–23). Sp inge Be lin Heidelbe g. h ps://doi.o g/10.1007/978-3-642-28616-2_1 Copeland, B. (2024). a i icial in elligence. Encyclopedia B i annica. h ps://www.b i annica.com/ echnology/a i icial-in elligence McDonough, M. (2024). la ge language model. Encyclopedia B i annica. Cha GPT4. (2024). Cha GPT. h ps://cha gp .com/ Amazon Web Se ices. (2024). Wha is GPT? Amazon. h ps://aws.amazon.com/wha -is/gp / 28 APPENDIX A - ETHICS COMMITTEE REPORT 29 APPENDIX B – DECISION ENGINE PROTOTYPE S age o BPM adop ion Focus poin s and sugges ions ele an o he s age Po en ial p oblems a he espec i e s age CSFs o add ess and me hods o use P oposed Sco ing app oach Measu men c i e ia Scale Da a sou ces Example o insigh and ecommenda ions Decision logic Empowe men o Employees: O e aining and wo kshops o imp o e employees' unde s anding o BPM and i s bene i s, and c ea e a sugges ion sys em ha enables hem o con ibu e ideas. Emplo yee Empowe men : Sco ing could be based on employee aining ini ia i es, he p esence o a sugges ions p og am, and he le el o au onomy employees ha e in hei oles. Numbe o BPM- ela ed aining sessions held. Employee pa icipa ion a e in BPM aining and wo kshops. Scale: 1 (No in ol emen ) o 5 (Full in ol emen and empowe men in BPM ac i i ies). In e nal sugges ion sys ems, HR eco ds. Ac ionable Insigh : Employees a e no sugges ing p ocess imp o emen s, indica ing a lack o empowe men . Recommenda ion: De elop and implemen an idea managemen sys em o cap u e employee sugges ions. Hold egula wo kshops and aining sessions o educa e employees on BPM p inciples. Encou age manage s o os e a cul u e ha alues inpu om all eam membe s. I he a e age sco e o employee empowe men is below 3, he decision engine ecommends p io i izing employee engagemen ini ia i es. Cus ome Focus: Implemen a eedback loop wi h cus ome s o unde s and hei needs and expec a ions, and in eg a e his eedback in o p ocess imp o emen plans. Cus ome Focus: Measu e he ex en o which cus ome eedback in luences p ocess changes and he alignmen o p ocesses o cus ome needs. Exis ence and u iliza ion o cus ome eedback mechanisms in p ocesses. Numbe o p ocess imp o emen s d i en by cus ome eedback. 1 (No in eg a ed) o 5 (Fully in eg a ed and d i ing BPM ini ia i es). Cus ome se ice eedback, p ocess documen a ion. I cus ome ocus is a ed high bu change openness is low, sugges s a egies o mo e esponsi e p ocess imp o emen s based on cus ome eedback. Openness o Changes: Encou age a cul u e ha alues adap abili y h ough egula communica ion abou he bene i s o change and ecogni ion o eams ha emb ace and lead success ul change ini ia i es. Openness o Change: Assess he o ganiza ion's his o y and eadiness o change, he exis ence o change managemen s a egies, and lexibili y in cu en p ocesses. Exis ence and u iliza ion o cus ome eedback mechanisms in p ocesses. Numbe o p ocess imp o emen s d i en by cus ome eedback. 1 (Resis an o any change) o 5 (Ac i ely emb aces change). Chang e log s, p ojec managemen ools. In ol emen and Full Suppo o Top Managemen : Ensu e ha op managemen no only endo ses bu is also ac i ely in ol ed in BPM ini ia i es by leading p ojec kick-o s and egula ly e iewing p og ess. Top Managemen In o l emen : Sco e based on isible suppo om leade ship, alloca ion o esou ces o BPM, and communica ion abou he impo ance o BPM om op managemen . F equency and consis ency o communica ion om op managemen ega ding BPM. In es men in BPM ini ia i es (budge , esou ces). 1 (No suppo ) o 5 (Highly isible and commi ed suppo ). Mee ing minu es, execu i e communica ions. Ac ionable Insigh : Top managemen in ol emen in BPM ini ia i es is minimal. Recommenda ion: Schedule egula BPM b ie ings o op managemen o epo on p og ess and alue. Appoin a senio execu i e as a BPM sponso o inc ease isibili y and commi men . C ea e a BPM s ee ing commi ee including op managemen o o e see ini ia i es. I op managemen suppo sco es less han 3, igge a ecommenda ion o execu i e wo kshops on BPM alue. I a p ojec champion is no iden i ied o sco es low, ecommend es ablishing a clea champion ole wi h de ined esponsibili ies and au ho i y. P ojec Champion: Appoin a espec ed leade wi hin he o ganiza ion o ac as a BPM champion who will ad oca e o and d i e BPM e o s. P ojec Champion P esence: E alua e he appoin men o p ojec champions o BPM ini ia i es and he au ho i y and esou ces gi en o hem. Iden i ica ion and ac i i y le el o BPM champions o p ocess owne s. Impac assessmen o p ojec champions on BPM p ojec s (p og ess, success s o ies). 1 (No champion) o 5 (S ong, e ec i e champion p omo ing BPM). P ojec documen a ion, s akeholde su eys. Business D i e s (a Sense o U gency): Iden i y and communica e he business impe a i es ha necessi a e BPM, such as ma ke p essu e o e iciency gains, o c ea e a sense o u gency o change. Sense o U gency/Business D i e s: De e mine whe he he e's a clea unde s anding ac oss he o ganiza ion o he need o BPM, d i en by compe i i e p essu e, e iciency goals, o o he business impe a i es. Documen a ion and communica ion o business impe a i es d i ing BPM adop ion. Response ime o c i ical p ocess- ela ed issues o ma ke changes. 1 (No d i e s iden i ied) o 5 (Clea d i e s iden i ied and communica ed wi h u gency). S a egic plans, in e nal communica ions. Clea ly De ined Objec i es and Plan: De elop a o mal p ojec cha e o each BPM p ojec ha ou lines i s objec i es, scope, expec ed bene i s, and imeline. P ojec Communica ion: Sco e he cla i y and equency o communica ion abou BPM p ojec goals, p og ess, and ou comes. Cla i y and dissemina ion o p ojec goals and miles ones ac oss he o ganiza ion. Numbe o documen ed and communica ed BPM p ojec plans. 1 (Ve y unclea ) o 5 (Ex emely clea and well unde s ood by all s akeholde s). P ojec cha e s, communica ion eco ds. Ac ionable Insigh : P ojec objec i es and plans a e no well-communica ed ac oss he o ganiza ion. Recommenda ion: U ilize a cen alized communica ion pla o m o dissemina e BPM p ojec upda es. Include clea communica ion plans in all BPM p ojec cha e s. Es ablish egula all-hands mee ings o discuss BPM p ojec s and in i e eedback. When sco es o p ojec communica ion and cla i y a e low, he logic emphasizes imp o ing p ojec cha e s and communica ion plans. I mo i a ion o change is high bu guidance om consul an s is low, he logic sugges s a e iew o consul an selec ion c i e ia and engagemen e ec i eness. P o essional Guidance o Ex e nal Consul an s: Engage wi h consul an s who ha e a p o en ack eco d in BPM o p o ide expe ise and an ex e nal pe spec i e. Ex e nal Consul an s' Guidance: Ra e he ex en and e ec i eness o ex e nal consul ancy use, including he ans e o knowledge and p ac ices o in e nal eams. Consul an engagemen sco e based on p ojec ou comes and eedback. ROI om con sul an -led ini ia i es e sus in e nally led. 1 (No in ol ed/no con ibu ion) o 5 (Signi ican in ol emen /con ib u ion). Consul an pe o mance e alua ions, p ojec ou come analyses. People Who A e Willing and Mo i a ed o Change: Selec p ojec eam membe s who a e open o change and ha e demons a ed adap abili y; in ol e hem in decision-making o keep hem mo i a ed. Change Willingness: Measu e he o ganiza ion's eadiness o change speci ic o BPM p ojec s, including s a engagemen and pa icipa ion. P e and pos - aining assessmen sco es on BPM eadiness. Employee u no e a es in depa men s unde going BPM p ojec s. 1 (High esis ance) o 5 (Eage willingness and mo i a ion o change). Engagemen su eys, pa icipa ion eco ds in BPM ini ia i es. Miscommunica ion: I no e ec i ely communica ed, BPM can be seen as a buzzwo d a he han a aluable me hodology, leading o skep icism. Insu icien Execu i e Suppo : Wi h ou s ong leade ship endo semen , BPM e o s may no be aken se iously o p io i ized. O e whelm wi h In o ma ion: An abundance o case s udies and in o ma ion can lead o con usion abou whe e o s a o un ealis ic expec a ions. Cul u al Resis ance: E en wi h awa eness, he e can be inhe en esis ance o change wi hin he o ganiza ion's cul u e ha is ha d o o e come. Misalignmen o Expec a ions: S akeholde s may ha e di e en expec a ions om BPM, leading o con lic s o disillusionmen i no managed well. Selec ion o Inco ec P ojec s: Choosing he w ong p ojec s o ini ial BPM e o s can esul in a lack o isible success and diminished c edibili y. P ojec Managemen Issues: Wi h ou s ong p o jec managemen , BPM p ojec s may ace delays, budge o e uns, o scope c eep. Inadequa e Measu emen : Poo ly de ined KPIs o a lack o measu emen ools can lead o ine ec i e moni o ing and he inabili y o demons a e alue. Desi e o Adop BPM Desc ip ion: The o ganiza ion acknowledges he need o BPM and is mo i a ed o s a adop ing BPM p ac ices. The e migh be some in o mal p ocess imp o emen s, bu no o ganized e o ye . The o ganiza ion decides o commi o BPM p ac ices and s a s o build in e nal desi e o change and p ocess imp o emen . Iden i ica ion: I he e is en husiasm abou BPM and discussions abou i s po en ial bene i s a e occu ing, bu no s uc u ed app oach has been aken, his indica es s age 2. Mo ing Fo wa d: De elop a clea ision and s a egy o BPM and s a engaging s akeholde s o suppo BPM ini ia i es. Focus: Building in e nal buy-in and p epa ing he o ganiza ion o change. Cul u e and Change Managemen : Fos e a cul u e ecep i e o change by in ol ing employees in he BPM discussion ea ly on. Add ess esis ance p oac i ely. S akeholde Engagemen : Iden i y and engage key s akeholde s o c ea e a coali ion ha desi es BPM adop ion, ensu ing hei needs and conce ns a e add essed. BPM P ojec s Desc ip ion: The o ganiza ion has ini ia ed speci ic BPM p ojec s, ocusing on imp o ing pa icula p ocesses. These a e o en pilo p ojec s ha se e as p oo o concep .Speci ic BPM p ojec s a e ini ia ed, o en as pilo s o ini ial e o s o apply BPM p inciples o pa icula business p ocesses. Iden i ica ion: I he e a e one o mo e BPM p ojec s unde way wi h dedica ed esou ces and goals, he o ganiza ion is in his s age. Mo ing Fo wa d: Analyze he ou comes o hese p ojec s o e ine he BPM app oach and p epa e o b oade implemen a ion. Focus: Gaining expe ience in BPM and demons a ing ea ly wins. Selec Pilo P ojec s: Choose ini ial BPM p ojec s ca e ully, a ge ing high-impac a eas whe e quick wins a e achie able. P ojec Managemen Bes P ac ices: U ilize s uc u ed me hodologies and obus p ojec managemen p ac ices o ensu e he success o BPM p ojec s. Moni o ing and Measu emen : Implemen KPIs o measu e he pe o mance o BPM p ojec s, using he da a o d i e decision- making and con inuous imp o emen . Awa eness and Unde s anding o BPM Desc ip ion: O ganiza ions a his s age ha e begun o ecognize he impo ance o BPM. P ocess imp o emen is ypically ad hoc, eac i e, and uns uc u ed. O ganiza ion ecognizes BPM as a po en ial alue-add and begins o unde s and i s p inciples and bene i s. Iden i ica ion: I BPM is a new concep o i he e is limi ed unde s anding o p ocess managemen p inciples o ganiza ion- wide, i 's likely a his s age. Mo ing Fo wa d: In oduce o mal BPM aining and es ablish a sense o u gency o p ocess imp o emen . Focus: Educa ion on BPM impo ance and i s impac on pe o mance. Educa ion a nd Co mmunica ion: Conduc wo kshops and aining sessions o educa e s akeholde s abou he bene i s and p inciples o BPM. Visible Execu i e Suppo : Secu e and showcase execu i e endo semen o highligh he impo ance o BPM wi hin he o ganiza ion. Case S udies and Success S o ies: Sha e case s udies o success ul BPM implemen a ions o illus a e po en ial gains and encou age buy-in. S age #1 S age #2 S age #3 30 APPENDIX C - EXPERT QUESTIONNAIRE 1. In you expe ience, wha a e he key elemen s o ocus on du ing he BPM adop ion? 2. How would you desc ibe he mos e ec i e me hods o in oduce BPM concep s o he o ganiza ion? 3. Wha ole do you see o leade ship in BPM adop ion? 4. Do you expec any challenges du ing implemen a ion o BPM? How can hey be mi iga ed? 5. Wha a e he mos c i ical success ac o s o BPM adop ion ha you belie e should be p io i ized? 6. I you can apply a decision engine model o BPM, wha ea u es do you hink would help you he mos ? Wha o a oid and how o s uc u e be e ? 7. How do you iew he po en ial ole o AI in BPM, especially in iden i ying and edesigning business p ocesses? 8. F om you pe spec i e, wha ypes o da a sou ces can be used o he decision engine? Would ex e nal inpu s like cus ome eedback add alue? 9. Should a BPM ool balance da a-d i en insigh s wi h human judgmen ? O be pu ely echnical? 31