Wo kshop
Recommended ci a ion: Ve kuilen, A. (2025). C ea ing AI Pe sonas o
Educa ional Se ings: A P epa ed4ed App oach. In Kangaslampi, R., Langie, G.,
Jä inen, H.-M., & Nagy, B. (Eds.), SEFI 53 d Annual Con e ence. Eu opean
Socie y o Enginee ing Educa ion (SEFI), Tampe e, Finland. DOI:
10.5281/zenodo.17631321.
This Con e ence Pape is b ough o you o open access by he 53 d Annual Con e ence
o he Eu opean Socie y o Enginee ing Educa ion (SEFI) a Tampe e Uni e si y in
Tampe e, Finland. This wo k is licensed unde a C ea i e Commons
A ibu ion-NonComme cial-Sha e Alike 4.0 In e na ional License.
CREATING AI PERSONAS FOR EDUCATIONAL SETTINGS: A
PREPARED4ED APPROACH
A. Ve kuilena
a The Hague Uni e si y o Applied Science, Del , The Ne he lands, , 0009-0004-
4146-2129
Con e ence Key A eas: Teaching and Lea ning Me hods, Digi al T ans o ma ion
Keywo ds: AI Pe sonas, Educa ional Technology, Pedagogical F amewo k,
Pe sonalized Lea ning, Responsible AI Use
ABSTRACT
This hands-on wo kshop equips enginee ing educa o s wi h p ac ical skills o design
and implemen AI pe sonas ha enhance eaching and lea ning expe iences. Using
he e idence-in o med PREPARED4ED amewo k, pa icipan s will lea n o c a
e ec i e p omp s ha align AI capabili ies wi h sound pedagogical p inciples.
Th ough li e demons a ions and guided exe cises, pa icipan s will expe ience how
AI pe sonas can suppo di e se educa ional goals— om guiding s uden s h ough
complex p oblem-sol ing o p o iding pe sonalized eedback while main aining
academic in eg i y. Wo king in collabo a i e g oups, a endees will c ea e and es
hei own AI pe sonas, i e a i ely e ining hem o add ess speci ic educa ional
challenges. The wo kshop emphasizes esponsible AI in eg a ion ha augmen s
a he han eplaces c i ical hinking, ensu ing ha echnology enhances a he han
diminishes au hen ic lea ning. Pa icipan s will lea e wi h wo king p o o ypes,
p ac ical implemen a ion s a egies, and a deepe unde s anding o how AI can be
e hically le e aged o ans o m enginee ing educa ion. This wo kshop o e s
aluable insigh s o all educa o s, ega dless o p io AI expe ience, who seek o
esponsibly inco po a e gene a i e AI in o hei eaching p ac ice.
1 BACKGROUND AND RATIONALE
1.1 The Eme ging Role o AI in Enginee ing Educa ion
Enginee ing educa ion s ands a a c i ical in lec ion poin as gene a i e AI
echnologies apidly ans o m educa ional landscapes. While hese ools o e
unp eceden ed oppo uni ies o pe sonaliza ion, immedia e eedback, and
in e ac i e lea ning expe iences, hey simul aneously p esen challenges o
adi ional eaching me hods, assessmen p ac ices, and academic in eg i y (Kasneci
e al., 2023; Roll & Wylie, 2016). Enginee ing educa o s inc easingly need o
na iga e his complex e ain, de e mining how o meaning ully in eg a e AI while
p ese ing he essen ial lea ning p ocesses ha de elop c i ical hinking and
p oblem-sol ing skills (Bies a, 2015).
Resea ch indica es a g owing challenge o di e en ia ing be ween human and AI-
gene a ed ex , emphasizing he u gen need o educa o s and s uden s o de elop
new compe encies and c i ical hinking skills o e ec i ely in eg a ing LLMs in
educa ion, while acknowledging conce ns abou po en ial misuse and p oposing
de ec ion me hods o AI-gene a ed con en alongside s a egies ha le e age hese
ools o enhance p oblem-sol ing abili ies, pa icula ly o olde lea ne s (Kasneci e
al., 2023). This c ea es an u gen need o educa o s o de elop expe ise no only in
using hese ools hemsel es bu in designing educa ional expe iences ha le e age
AI's s eng hs while mi iga ing po en ial isks such as cogni i e o loading,
o e eliance on echnology, and su ace-le el unde s anding (Zawacki-Rich e e al.,
2019).
1.2 AI Pe sonas as Pedagogical Tools
AI pe sonas—cus omized AI in e aces designed wi h speci ic pedagogical oles and
cha ac e is ics— ep esen a p omising app oach o in eg a ing AI in o educa ional
se ings. When hough ully de eloped, hese pe sonas can se e as lea ning
companions, guides, o o ma i e assessmen ools ha align wi h speci ic
educa ional objec i es (Khan, 2023). They o e a s uc u ed way o ha ness AI
capabili ies while main aining clea pedagogical in en ionali y (Ve kuilen & G i ioen,
2024).
Fo enginee ing educa ion speci ically, AI sy ems, like AI pe sonas, can help add ess
pe sis en challenges such as p o iding indi idualized guidance in la ge classes,
o e ing jus -in- ime suppo o complex p oblem-sol ing, and c ea ing oppo uni ies
o s uden s o a icula e and es hei unde s anding h ough dialogue (Roll & Wylie,
2016). Howe e , c ea ing e ec i e AI pe sonas equi es mo e han echnical
knowledge o p omp ing echniques; i demands an unde s anding o how o align AI
capabili ies wi h sound educa ional p inciples (Tomlinson, 2014).
Lo em ipsum dolo si ame , consec e u adipiscing eli , sed do eiusmod empo
incididun u labo e e dolo e magna aliqua. U enim ad minim eniam, quis nos ud
exe ci a ion ullamco labo is nisi u aliquip ex ea commodo consequa . Duis au e i u e
dolo in ep ehende i in olup a e eli esse cillum dolo e eu ugia nulla pa ia u .
Excep eu sin occaeca cupida a non p oiden , sun in culpa qui o icia dese un
molli anim id es labo um.
1.3 The PREPARED4ED F amewo k
AI pe sonas need o be in en ionally ained as eache s, no jus as sou ces o
in o ma ion, because—as Wiley (2023) a gues— eaching is a dis inc p o essional
ac i i y, no a side hus le. The PREPARED4ED amewo k o e s a s uc u ed,
e idence-in o med app oach o designing AI educa ional expe iences (Ve kuilen &
G i ioen, 2024). G ounded in es ablished educa ional heo ies including Bloom's
Taxonomy, Sel -De e mina ion Theo y (Ryan & Deci, 2000), Vygo sky's Zone o
P oximal De elopmen (Vygo sky, 1978), and Kolb's Expe ien ial Lea ning Cycle
(Kolb, 1984), his amewo k guides educa o s h ough he p ocess o c ea ing AI
pe sonas ha se e meaning ul educa ional pu poses.
Each elemen o he PREPARED4ED ac onym add esses a c i ical aspec o
e ec i e AI in eg a ion:
• Pu pose ul P omp : Clea alignmen wi h speci ic lea ning objec i es
• Rele an Role: Connec ed o s uden s' in e es s and p io knowledge
• Expe ien ial and Explici : In ol ing au hen ic applica ion and p ac ice
• P og essi e and Ask: Building complexi y g adually o manage cogni i e
load (Swelle e al., 1998)
• Assessable and Pa ame e s: Suppo ing meaning ul e alua ion o
lea ning
• Re lec i e: Encou aging me acogni ion and sel - egula ion
• Engaging Emo ion: S imula ing cu iosi y and c i ical hinking
• Di e en ia ed: Adap ing o di e se lea ning needs and p e e ences
(Tomlinson, 2014)
By applying his amewo k o he de elopmen o AI pe sonas, educa o s can
ensu e ha hei use o AI echnology is pedagogically sound, e hically esponsible
(P insloo & Slade, 2017), and genuinely bene icial o s uden lea ning.
This wo kshop add esses a c i ical gap in enginee ing educa ion by p o iding
educa o s wi h bo h he concep ual unde s anding and p ac ical skills needed o
de elop e ec i e AI pe sonas ha enhance a he han diminish au hen ic lea ning
expe iences (Bies a, 2015; Khan, 2023).
2 WORKSHOP OBJECTIVES
2.1 Ta ge audience
This wo kshop is designed o enginee ing educa o s ac oss a ious oles who seek
o e ec i ely in eg a e AI in o hei eaching p ac ice:
• Lec u e s and p o esso s looking o inno a e hei eaching app oaches
• Labo a o y ins uc o s seeking in e ac i e ools o hands-on lea ning
• Cu iculum designe s in e es ed in inco po a ing AI-enhanced lea ning
expe iences
• Educa ional de elope s suppo ing acul y in echnology in eg a ion
• P og am di ec o s conside ing he s a egic implemen a ion o AI ac oss
cou ses
• The wo kshop is accessible o pa icipan s wi h a ying le els o AI
expe ience:
• Newcome s o AI will gain ounda ional unde s anding and p ac ical
s a ing poin s
• Educa o s wi h basic AI amilia i y will lea n s uc u ed app oaches o
enhance hei p ac ice
• Expe ienced AI use s will bene i om he pedagogical amewo k o e ine
hei implemen a ions
No specialized echnical knowledge is equi ed beyond basic digi al li e acy.
Pa icipan s should b ing a lap op o able wi h in e ne access and ha e o c ea e a
( ee) Cha GPT accoun be o e a ending. A bi o p omp ing knowledge is help ul bu
no necessa y.
2.2 Expec ed lea ning ou comes
By he end o his wo kshop, pa icipan s will be able o:
1. Apply he PREPARED4ED amewo k o design pedagogically sound AI
pe sonas o speci ic educa ional con ex s in enginee ing
2. C ea e e ec i e p omp s ha consis en ly gene a e help ul, accu a e
esponses aligned wi h in ended lea ning ou comes
3. C i ically e alua e AI pe sonas o po en ial issues including cogni i e
o loading, bias, and academic in eg i y conce ns
4. Implemen s a egies o using AI pe sonas ha enhance s uden
engagemen and ac i e lea ning a he han eplacing essen ial cogni i e
p ocesses
5. Design di e en ia ed lea ning expe iences using AI pe sonas ha add ess
di e se s uden needs and lea ning p e e ences
6. De elop implemen a ion plans o in eg a ing AI pe sonas in o exis ing
cou ses while main aining pedagogical in eg i y
7. A icula e p inciples o esponsible AI use in enginee ing educa ion o
colleagues and s uden s
These ou comes e lec a balanced app oach ha emb aces inno a ion while
main aining c i ical awa eness o bo h he po en ial and limi a ions o AI in
educa ional se ings.
3 WORKSHOP DESIGN
3.1 Time plan
This wo kshop combines sho p esen a ions, li e demons a ions, collabo a i e
g oup wo k, and pee eedback. You'll mo e be ween explo ing ideas, c ea ing AI
pe sonas, es ing hem, and e ining you designs. The mix o ac i i ies is designed
o keep hings p ac ical, in e ac i e, and di ec ly ele an o you eaching.
Table 1. Wo kshop ime plan
Run ime
Ac i i y
Desc ip ion
En y
minu es
Fo ma ion o wo king g oups
Pa icipan s g oup acco ding o
educa ional challenges/in e es s.
Simply by si ing a he able wi h he AI
pe sona(s) o hei choice.
5 minu es
In oduc ion o wo kshop
Wha , when and wha no . O e iew o
AI and i s educa ional implica ions
5 minu es
You jou ney wi h AI?
Ge ing o know each o he and hei
jou ney wi h AI
10 minu es
Video exe cise p omp ing
P omp ing needs o be speci ic, which
elemen s should i con ain?
5 minu es
P omp you own AI Pe sona
De elop ini ial AI pe sonas based on
he p omp amewo k Role, Goal,
Fo ma & Wo k low
5 minu es
In oducing PREPARED4ED
F amewo k
Explana ion o amewo k elemen s wi h
enginee ing educa ion examples and a
demons a ion o he F amewo k a
wo k wi h he PREPARED4ED GPT
20 minu es
Designing AI Pe sonas
Wo king wi h he PREPARED4ED GPT
o imp o e he p omp o he AI
Pe sona. Tes ing he AI Pe sona.
10 minu es
Lessons Lea ned
Sha ing ob ained esul s and
expe iences
3.2 In e ac i i y
The wo kshop is designed as a hands-on, collabo a i e expe ience wi h a a ie y o
ac i i y ypes o suppo ac i e lea ning and p o essional dialogue:
1. Sel -selec ion and pee connec ion: Pa icipan s begin by g ouping
hemsel es based on in e es in speci ic AI pe sonas, os e ing immedia e
ele ance and pee alignmen . Be o e he wo kshop s a s his will al eady
c ea e discussions abou wha is impo an in hei educa ional se ings.
2. Dialogic and expe ien ial lea ning: Sho plena y sessions in oduce co e
concep s and in i e pa icipan s o sha e hei own AI jou neys, building on
p io knowledge and es ablishing a sha ed ounda ion.
3. In e ac i e p omp ing exe cise based on a ideo: A guided ac i i y using
ideo helps pa icipan s o unde s and ha p omp ing is no easy and o
iden i y wha makes a p omp e ec i e, se ing he s age o pe sona
c ea ion.
4. Hands-on pe sona design wi h sca olded suppo : Using he Role, Goal,
Fo ma & Wo k low amewo k ollowed by he PREPARED4ED
amewo k, pa icipan s c ea e and i e a i ely e ine AI pe sonas using
GPT ools, pee and acili a o eedback. G oups will es hei AI pe sonas
wi h eal que ies and scena ios, documen ing bo h success ul in e ac ions
and limi a ions. This i e a i e p ocess mi o s au hen ic design
me hodologies and p o ides immedia e e idence o e ec i eness.
5. Collabo a i e e lec ion and knowledge sha ing: G oups es hei
pe sonas, sha e insigh s, and su ace lessons lea ned in a closing ound
o consolida e lea ning and suppo ans e o p ac ice.
Th oughou hese ac i i ies, acili a o s will ci cula e among g oups, p o iding
guidance, answe ing ques ions, and helping pa icipan s o e come echnical o
concep ual challenges. The wo kshop emphasizes lea ning by doing, wi h b ie
ins uc ional segmen s ollowed by ex ended pe iods o ac i e engagemen wi h he
ools and concep s.
To ensu e maximum pa icipa ion and accessibili y, he wo kshop will:
• P o ide empla e documen s and examples o sca old he design p ocess
• Inco po a e bo h indi idual e lec ion and g oup discussion componen s
• Main ain a balance be ween minimal echnical skills and pedagogical
conside a ions
• Add ess e hical dimensions o AI use h oughou , a he han ea ing hem
as a sepa a e opic, bu no explici since ha is no he main opic o he
wo kshop.
This mul i ace ed app oach ensu es ha pa icipan s emain ac i ely engaged
h oughou he session while de eloping p ac ical skills hey can immedia ely apply
in hei own educa ional con ex s.
4 WORKSHOP RESULTS
[This sec ion will be comple ed a e he con e ence wi h indings, pa icipan
eedback, and ou comes o he wo kshop.]
REFERENCES
Bies a, G. (2015). Wha is educa ion o ? On good educa ion, eache judgemen ,
and educa ional p o essionalism. Eu opean Jou nal o Educa ion, 50(1), 75-87.
h ps://doi.o g/10.1111/ejed.12109
Kasneci, E., Sessle , K., Küchemann, S., Banne , M., Demen ie a, D., Fische , F.,
Gasse , U., G oh, G., Günnemann, S., Hülle meie , E., K usche, S., Ku yniok, G.,
Michaeli, T., Ne del, C., P e e , J., Poque , O., Saile , M., Schmid , A., Seidel, T., ...
Kasneci, G. (2023). Cha GPT o good? On oppo uni ies and challenges o la ge
language models o educa ion. Lea ning and Indi idual Di e ences, 103, 102274.
h ps://doi.o g/10.1016/j.lindi .2023.102274
Khan, S. (2023, July 19). Reimagining educa ion in he age o AI [Video]. TED.
h ps://www. ed.com/ alks/sal_khan_ eimagining_educa ion_in_ he_age_o _ai
Kolb, D. A. (1984). Expe ien ial lea ning: Expe ience as he sou ce o lea ning and
de elopmen . P en ice-Hall.
P insloo, P., & Slade, S. (2017). E hics and lea ning analy ics: Cha ing he
(un)cha ed. In C. Lang, G. Siemens, A. F. Wise, & D. Gaše ic (Eds.), The handbook
o lea ning analy ics (pp. 49-57). Socie y o Lea ning Analy ics Resea ch.
h ps://www.sola esea ch.o g/publica ions/hla-17/hla17-chap e 4/
Roll, I., & Wylie, R. (2016). E olu ion and e olu ion in a i icial in elligence in
educa ion. In e na ional Jou nal o A i icial In elligence in Educa ion, 26(2), 582-599.
h ps://doi.o g/10.1007/s40593-016-0110-3
Ryan, R. M., & Deci, E. L. (2000). Sel -de e mina ion heo y and he acili a ion o
in insic mo i a ion, social de elopmen , and well-being. Ame ican Psychologis ,
55(1), 68-78. h ps://doi.o g/10.1037/0003-066X.55.1.68
Swelle , J., an Me iënboe , J. J. G., & Paas, F. (1998). Cogni i e a chi ec u e and
ins uc ional design. Educa ional Psychology Re iew, 10, 251-296.
h ps://doi.o g/10.1023/A:1022193728205
Tomlinson, C. A. (2014). The di e en ia ed class oom: Responding o he needs o
all lea ne s (2nd ed.). ASCD.
Ve kuilen, A., & G i ioen, E. (2024). P ac ical use o AI-pe sona's o pe sonalised
lea ning expe iences. Wo kshop p esen ed a Online Educa Be lin 2024.
Vygo sky, L. S. (1978). Mind in socie y: The de elopmen o highe psychological
p ocesses. Ha a d Uni e si y P ess.
Wiley, D. (2023, No embe 29). Making AI a Mo e E ec i e Teache : Lessons om
TPACK. OpenCon en . h ps://opencon en .o g/blog/a chi es/7705
Zawacki-Rich e , O., Ma ín, V. I., Bond, M., & Gou e neu , F. (2019). Sys ema ic
e iew o esea ch on a i icial in elligence applica ions in highe educa ion – whe e
a e he educa o s? In e na ional Jou nal o Educa ional Technology in Highe
Educa ion, 16(1), 39. h ps://doi.o g/10.1186/s41239-019-0171-0
AUTHOR NOTE:
The au ho used gene a i e AI ools o e ine aspec s o he ex o imp o ed cla i y
and eadabili y. All con en , accu acy, and scien i ic igo emain he sole
esponsibili y o he au ho .
