Pa e ns and Themes in RCD P o essional Ca ee Pa hs
Samuel Po e
Uni e si y o Ma yland
[email p o ec ed]
Pa ick Schmi z
Sempe Cogi o
[email p o ec ed]
Deb McCa ey
A izona S a e Uni e si y
[email p o ec ed]
Ma hew Humph eys
A izona S a e Uni e si y
[email p o ec ed]
Ke k Kee
Texas Tech Uni e si y
[email p o ec ed]
Abs ac
Resea ch Compu ing and Da a (RCD) p o essionals play a c ucial ole in suppo ing and ad anc-
ing esea ch ha in ol es da a and/o compu ing, howe e , he e is a c i ical sho age o hese RCD
p o essionals and o ganiza ions ace challenges in ec ui ing and e aining RCD wo k o ce oles. Many
po en ial RCD p o essionals a e unawa e o hese oles and how hei skills and expe ience map o he
RCD wo k o ce needs. The CaRCC RCD Ca ee A cs wo king g oup has been wo king o gain a deepe
unde s anding o RCD ca ee pa hs. Following up on ou su ey o key ac o s ha in luence decisions
in he cou se o hei ca ee s, we conduc ed 14 in e iews o mo e deeply explo e ca ee choices, in lu-
ences, and ac o s ha help shape RCD ca ee pa hs. This pape summa izes he indings and p esen s
obse a ions and ecommenda ions o bo h hi ing manage s as well as cu en RCD p o essionals.
Keywo ds— Resea ch Compu ing and Da a, RCD, Cybe in as uc u e, CI, Ca ee Pa hs, P o essionaliza ion
1 In oduc ion
“I didn’ know hese jobs exis ed.” This is a quo e om a ecen candida e who applied o a Resea ch Compu ing
and Da a (RCD) ole a one o ou ins i u ions, and i echoes he sen imen o en sha ed by colleagues: hey ne e
expec ed o become an RCD p o essional – hey didn’ know hese oles exis ed in he i s place! Finding people o ill
RCD oles a all le els con inues o be a challenge o manage s. While e minology and HR amewo ks o hese oles
a e being adop ed[4], he e is s ill much p og ess o be made in e ms o unde s anding how bes o ec ui , e ain,
and suppo RCD p o essionals[1]. RCD in as uc u e and se ices a e essen ial o he mode n esea ch en e p ise
and con inue o e ol e and expand a a apid pace. Resea che s a e s uggling o keep pace wi h he explosion o
da a and he apid e olu ion o compu ing esou ces and o en lack he skills o make ull use o eme ging ools
and echniques. RCD p o essionals p o ide he suppo esea che s need o ake ad an age o he la es ools and
se ices [11]. The CaRCC Ca ee A cs Wo king G oup was o med o in es iga e RCD ca ee pa hs and ac o s
in luencing ca ee decisions.
Se e al wo kshops ha e iden i ied he impo ance o unde s anding RCD ca ee pa hs. The 2017 NSF-sponso ed
P o essionaliza ion in Cybe in as uc u e wo kshop [2] emphasized he sca ci y o RCD p o essionals. The indings
o a 2020 wo kshop, Building he Resea ch Inno a ion Wo k o ce [10] emphasized “... he e a e ew well de ined and
s able ca ee pa hs” and “...i is no clea wha ca ee pa hs a e a ailable o how a pe son can pu sue ad ancemen .”
A b oad su ey o RCD s akeholde s[6] ound ha he issue “Documen ing and suppo ing aspi a ional ca ee pa hs
o RCD p o essionals wi hin highe educa ion” was a ed as bo h e y impo an and qui e di icul .
Resea ch by he CaRCC RCD P o essionaliza ion Wo king G oup[8] demons a ed he wide ange o s a ing poin s
and ul ima e oles o RCD p o essionals. They also no e he need o clea ca ee pa hs and a be e unde s anding
o how o e ain and de elop RCD p o essionals. O he esea ch e o s ha e iden i ied he impo ance o men o ing
and p o essional de elopmen [3, 9]. Filinge e al.[7] discuss he impo ance o lea ning pa hways and he need o help
ea ly-ca ee RCD p o essionals unde s and he oles and he needed skills o en e RCD ca ee s and be success ul.
Hacke e al.[10] also no ed he lack o ole models and men o s wi hin he communi y.
The Ca ee A cs Wo king G oup’s ea lie wo k[5] explo ed he key ac o s ha RCD P o essionals conside ed in
making ca ee decisions. Ou cu en wo k ollows up on ha wi h 14 in e iews o RCD p o essionals a a ious ca ee
s ages. Analysis o he esul ing da a u he explo ed commonali ies in how RCD p o essionals ound hese oles,
how hey g ew in o hem, and wha ac o s and in luences impac ed hei ca ee s o e ime. This pape summa izes
ou app oach o his wo k, wha we lea ned, and some ecommenda ions o bo h po en ial RCD p o essionals as well
as manage s hi ing o hese oles.
©2025 CaRCC licensed o use unde C ea i e Commons A ibu ion-NoDe i a i es 4.0 In e na ional (CC BY-ND 4.0) license.
2 Me hodology
This quali a i e s udy used semi-s uc u ed in e iews wi h 14 di e se RCD p o essionals o explo e hei ca ee
ajec o ies. Conduc ed ia Zoom be ween May 2023 and Feb ua y 2024, he in e iews examined educa ional and
p o essional backg ounds, pi o al ca ee momen s, mo i a ions, challenges, and u u e aspi a ions. G ounded heo y
me hodology[12] guided he analysis h ough open, axial, and selec i e coding, g ouping hemes in o h ee ca ee
s ages (p e/ea ly, mid, and ad anced) and ou cen al opics (decisions made, mo i a ions, ac o s and in luences,
and isks aken). Pa icipan s’ demog aphics a ied in gende , ace/e hnici y, age, and p o essional backg ounds,
wi h mos employed in academia o na ional labs. The s udy eached heo e ical sa u a ion and p oduced a ca ee
ajec o y model consis en wi h p io esea ch, o e ing ich insigh s in o how RCD ca ee s e ol e h ough key
decisions, mo i a ions, and isk- aking beha io s.
3 Wha We Lea ned
3.1 En e ing he RCD Field
The s o ies o hose RCD p o essionals who began on an academic ca ee pa h demons a ed ha many en e ed RCD
as a unc ion o oppo uni y. Mos did no in en ionally seek ou hei i s RCD ole, bu we e in oduced o i
by a men o , o ound i while looking ou side o academia due o a lack o a ac i e academic oppo uni ies (e.g.,
hey did no look o wa d o he pu sui o g an unding and enu e compe i ion bu wan ed o emain engaged in
esea ch). One said: “I s umbled in o someone who showed me a posi ion in esea ch compu ing ha I el I could do
while looking o mo e ull- ime employmen . I ound ha I liked i a lo !”
Some RCD p o essionals ansi ioned om indus y oles, ech s a ups, manu ac u ing, en e p ise IT (including
campus IT), o co po a e esea ch labs. These ca ee shi s we e mo i a ed by changing alues (no suppo ing he
mission o hei company anymo e), a desi e o con ibu e posi i ely o socie y, o a desi e o a “mo e elaxed” wo k
en i onmen .
While a PhD is no equi ed o mos RCD oles, many pa icipan s epo ed ha ha ing a PhD and/o expe ience
wi h esea ch, including a ack eco d o publica ions, helped in in e ac ions wi h esea che s, who app ecia ed hei
unde s anding o he esea ch me hods. While expe iences a y by ins i u ion and ole, many people in RCD oles
a e equally success ul wi hou a PhD.
3.2 Impo ance o Men o ing, Communi y, and Ne wo k
In ea ly- and mid-ca ee s ages, men o s (bo h o mal and in o mal) and communi ies we e desc ibed as signi ican
ac o s in RCD ca ee pa hs. By he same oken, ad anced-ca ee pa icipan s exp essed a desi e o men o and coach
o he s. O en, men o s o supe iso s we e esponsible o in oducing pa icipan s o he RCD ield. Pa icipa ion
in communi ies o p ac i ione s – whe he a hei ins i u ion, in hei egion, o na ionally – led o an inc ease in
con idence abou skills, c ea ing a sense o belonging, and impac ing ca ee decisions, including whe he o no o
con inue in he RCD ield. These indings unde sco e he impo ance o e o s in he RCD communi y o expand
men o ship and coaching se ices, including he new CaRCC Men o ing Wo king G oup1.
3.3 Mo i a ions, Fac o s, and Risk Tole ance a y by Ca ee S age
Compa ing ac o s and mo i a ions ha in luenced decisions ac oss ca ee s ages, we ound a majo shi be ween
ea ly-ca ee and mid-ca ee . Ea ly-ca ee decisions we e o en based on ac o s like inancial necessi y, ca ee pa h
dissa is ac ion, o a lack o oppo uni y o ad ancemen . These ac o s align wi h a gene al willingness o ole a e
mo e isk, e.g., changing ca ee pa hs, going back o school, o aking on oles wi h unce ain p ospec s. Howe e o
mid-ca ee p o essionals, ac o s shi owa d being in e es ed in he wo k, an abili y o ha e in luence, and sala y.
In bo h mid- and ad anced-ca ee s ages, pa icipan s desc ibed an inc eased ocus on pe sonal p io i ies, meaning
job changes we e mo i a ed by pe sonal goals and no necessi y. Mo i a ions o change included he igh wo k
en i onmen (some imes ge ing away om poli ics o a shi in mission), a desi able wo k loca ion, and desi able
ci cums ances o he ole, such as building a new eam e sus managing a well- unc ioning g oup. Ad anced-ca ee
ac o s included impac a a la ge scale, aking on new challenges, and sa is ac ion in coaching and men o ing. As
hey ad ance, people choose he ole hey wan , no he job hey need, al hough hese decisions a e o en balanced
by inc eased isk-a e sion due o amily and inancial cons ain s. O no e, when asked o imagine hei u u e sel es,
mos pa icipan s indica ed hey expec ed o emain in RCD oles.
1h ps://ca cc.o g/men o ing-wo king-g oup/
2
In a common mid-ca ee choice, we saw pa icipan s sel -selec in o o away om managemen oles e lec ing
hei e ol ing sel -awa eness o an inclina ion o manage o he s o o be an indi idual con ibu o . Some mid- and
ad anced-ca ee pa icipan s desc ibed eaching he end o a p omo ion pa h a hei ins i u ion, p omp ing a mo e
o ind new oppo uni ies. This is pa icula ly e iden in managemen and leade ship oles o which he e a e ewe
posi ions, bu also occu s in non-supe iso y oles when ins i u ions do no ha e well-de ined pa hs o ad ancemen .
A he same ime, o he pa icipan s desc ibed a s ong sense o loyal y and commi men o hei ins i u ion and said
hey wouldn’ lea e hei posi ion wi hou comple ing in-p og ess p ojec s.
3.4 Common T ai s
The RCD p o essionals we in e iewed a e passiona e abou some aspec s o hei ole bu we ound ha hei passions
shi ed o e ime, o en in line wi h shi ing oles. Fo example, one pa icipan exp essed a passion o p og amming
and compu e s ha kep hem in an RCD ole ea ly in hei ca ee (and li e ally in a lab). Now well in o mid-ca ee ,
a clea passion is coaching and de eloping people. O he examples we hea d pa icipan s alk abou we e how much
hey like a gi en domain (e.g., bioin o ma ics o chemis y), sol ing p oblems, leading eams, o p o iding a se ice
o esea che s. One pa icipan explained: “Ha ing a job whe e I ge o in e ac wi h esea ch and suppo i om a
se ice mindse is eally ewa ding.”
We ind ha hese ai s e lec a eali y ha RCD p o essionals o en ace when p o iding suppo o esea ch wi h
a signi ican amoun o scope- and esou ce-ambigui y. Pa icipan s desc ibed hemsel es as sel -di ec ed, passiona e,
cu ious lea ne s who a e willing o ask o help, sha e, and be open o new ideas. Though hese ai s don’ gua an ee
success o e lec all people who a e in RCD oles, hey ep esen a common h ead unning h ough many o he
in e iews.
3.5 Recommenda ions o Manage s
1. Conside p io i izing e o s o connec new RCD p o essionals o men o s and pee RCD p o essionals, and
encou age (and allow ime o ) in ol emen in RCD communi ies.
2. Recognize ha ea ly-ca ee esea che s a e o en aking signi ican isks o en e he RCD ield and p o ide
suppo o his change.
3. Manage s should engage hei ins i u ions along wi h he RCD communi y o be e de ine HR oles and ca ee
pa hs o make RCD pa hs seem less isky.
4. When ec ui ing and hi ing, communica e wha RCD p o essionals ind ewa ding abou hei oles and empha-
size ai s such as cu iosi y as much as o mo e han echnical skills.
5. Ea ly-ca ee p o essionals a e good candida es o RCD oles, and hi ing manage s should conside a ange o
ields/backg ounds. Manage s mus desc ibe ca ee pa hs, and may wan o connec p ospec s wi h cu en
RCD p o essionals who can ad oca e o RCD oles.
6. When ec ui ing and hi ing mid- and ad anced-ca ee RCD p o essionals, keep in mind ha hey may be mo e
isk-a e se and less willing o eloca e, bu will o en make la e al mo es o u he pe sonal goals. By he same
oken, employe s can e ain hese p o essionals by ocusing on wha hey ind ul illing.
3.6 Insigh s o New and Cu en RCD P o essionals
1. Pa icipan s om many ields and backg ounds ound success and ul illmen in an RCD ca ee , no jus p o-
essionals wi h a compu ing backg ound o an academic backg ound. No ably, a PhD is no equi ed, bu
p o essionals wi h a PhD o esea ch expe ience said hey el be e equipped o unde s and he esea che s
hey suppo and a e sa is ied by he wo k.
2. Pa icipan s equen ly desc ibed he impo ance o men o ship and hei in ol emen in communi ies o p ac ice
o hei ca ee de elopmen and enjoymen .
3. Many pa icipan s ad anced wi hin hei ins i u ions by aking on supe iso y oles, bu he e a e also op-
po uni ies o ad ance o non-supe iso s, such as echnical o domain expe ise (leade ship as dis inc om
managemen ). Some RCD p o essionals ound success h ough sel -ad ocacy, o by mo ing ac oss ins i u ions
in o oles ha i hei ca ee aspi a ions, pa icula ly when he oles a e no supe iso y in na u e.
4. This ield is new and many ins i u ions a e in es ing in hese oles o he i s ime. You can in luence you
own g ow h and may be able o de ine you own ole.
5. Ve y ew pa icipan s wan ed o lea e he RCD ield; people like hese jobs. Those ha le gene ally did so
due o changes in he wo k en i onmen and no because o he wo k i sel .
3
3.7 Visualizing RCD Ca ee Pa hs
In addi ion o he w i en analysis, his phase o wo k included he goal o p oducing a isualiza ion o help RCD
p o essionals unde s and he ca ee pa hs open o hem and o help manage s unde s and sou ces o hi ing and how
o suppo exis ing RCD p o essionals wi h hei p o essional de elopmen . We c ea ed a isualiza ion inspi ed by
ansi maps, and submi ed as a pos e o PEARC25 [in submission]; he ull image o he pos e is a ailable a :
doi.o g/10.5281/zenodo.15048236.
Figu e 1: Exce p o he RCD Ca ee Pa hs Visualiza ion.
4 Conclusions and Fu u e Wo k
We se ou o explo e ac o s ha a ec RCD ca ee pa hs, and o iden i y eme gen hemes a ound he in e ac ions
be ween hose ac o s. Despi e ou small sample size, ou wo k elucida ed much in se ice o hese goals. Ou
wo k indica es ha RCD ca ee pa hs a e la gely conside ed as al e na i e ca ee pa hs: ound by academic ca ee -
seeke s who de ia e om a adi ional academic ca ee pa h o indus y ca ee -seeke s who eel dissa is ied wi h hei
indus ial en i onmen . We also ound ha an indi idual’s ca ee s age s ongly a ec ed hei mo i a ions o aking
on new oles, and hei app oach o isk. A lack o awa eness and aspi a ion o RCD oles means manage s will
need o be hough ul when desc ibing, hi ing, and ec ui ing o RCD oles. Few ea ly-ca ee indi iduals will i
an RCD ole pe ec ly. O ganiza ions building RCD eams should alue cu iosi y, sel -di ec ion, and communica ion
among ea ly-ca ee oles, as hese quali ies seemed o align wi h pa icipan s’ ad ancemen and g ow h. Ou wo k
also demons a ed ha he e is a g ea deal o a iance in RCD ca ee pa hs. PhD’s p o ided people wi h meaning ul
ad an ages, and o hose who came om academia, o en led hem o he RCD ield in he i s place. Howe e , PhD’s
a e no equi ed o ha e a success ul and impac ul RCD ca ee . Ad ancemen ypically occu ed wi hin o ganiza ions,
ega dless o educa ion le el. Mo es o o he o ganiza ions we e usually la e al in na u e, hough hey opened up
u he ad ancemen and mo e pe sonally ul illing oles. Simila ly, men o ship was o en e y impo an in ea ly- and
mid-s age ca ee s due o i s impo ance in inding he ield, o de eloping capabili ies and quali ies ha indi iduals
may ha e lacked, o o i s impo ance o ad ancemen . Howe e , some indi iduals lacked good men o ship en i ely;
hei ca ee pa hs we e o en based on hei abili y o adap o shi ing o ganiza ional needs o en i onmen s.
Despi e hese p ojec successes, he e is s ill wo k o be done. Oppo uni ies include ga he ing a la ge sample
size, e ining coding p ocedu es, and aking a mo e quan i a i e app oach o analyze he empo ali y, di ec ionali y,
and s eng h o a ious ac o s. This would be aluable as a ious mode a ion and media ion hypo heses eme ged as
a esul o ou wo k. Addi ionally, hese would open up possible analysis h ough g aph heo y algo i hms.
4
Acknowledgemen s
This wo k has been suppo ed in pa by Na ional Science Founda ion g an s OAC-1620695 and OAC-2100003. Special
hanks a e owed o he communi y membe s who we e in e iewed o his wo k, and o ou o he CaRCC Ca ee
A cs wo king g oup membe s Gladys Andino, Sha aq Chaudh y, and P es on Smi h who con ibu ed o he RCD
Ca ee A cs e o .
Re e ences
[1] Lisa A a une, Dana B unson, Thomas Hacke , and P es on Smi h. Building he esea ch inno a ion wo k o ce:
a wo kshop o iden i y new insigh s and di ec ions o ad ance he esea ch compu ing communi y. Technical
epo , Feb ua y 2021.
[2] Nicholas Be en e, James Howison, Joel Cu che -Ge shen eld, John L. King, S ephen R. Ba ley, and John Towns.
P o essionaliza ion in Cybe in as uc u e. SSRN Schola ly Pape ID 3138592, Social Science Resea ch Ne wo k,
Roches e , NY, Feb ua y 2017.
[3] Sco Callaghan, Elsa Gonsio owski, We onika Filinge , and Ilya Zhuko . De eloping a Success ul HPC Men o ing
P og am a he IHPCSS. In P ac ice and Expe ience in Ad anced Resea ch Compu ing 2024: Human Powe ed
Compu ing, pages 1–9, P o idence RI USA, July 2024. ACM.
[4] CaRCC. Resea ch Compu ing and Da a P o essionals - Job Elemen s and Ca ee Guide, Ap il 2018.
[5] Sha aq Chaudh y, A man Pazouki, Pa ick Schmi z, Elizabe Hille y, and Ke k Kee. Unde s anding Fac o s ha
In luence Resea ch Compu ing and Da a Ca ee s. In P ac ice and Expe ience in Ad anced Resea ch Compu ing
(PEARC ’22), Bos on, MA, USA, July 2022. Associa ion o Compu ing Machine y.
[6] Joel Cu che -Ge shen eld, To ey Ba elle, Dana B unson, Thomas Chea ham, Jacob Fosso Tande, Douglas Jen-
newein, Julie Ma, Lau en A. Michael, Timo hy Middelkoop, Hen y Neeman, and Pa ick Schmi z. P o ession-
aliza ion o Resea ch Compu ing and Da a: An Expanded Agenda. In P ac ice and Expe ience in Ad anced
Resea ch Compu ing, pages 129–136, Po land OR USA, July 2023. ACM.
[7] We onika Filinge , Julie Mullen, Je emy Cohen, Saman ha Wi ke, and Ann Backhaus. Building HPC Lea ning
Pa hways: Unde s anding ou Communi y. In P ac ice and Expe ience in Ad anced Resea ch Compu ing 2024:
Human Powe ed Compu ing, pages 1–7, P o idence RI USA, July 2024. ACM.
[8] Ri ika Gi i, Sco Yockel, Kimbe ly G asch, Ashley S au e , S e an Robila, and Ch is Reidy. Examining cd
oles and ac i i ies h ough he lens o acings, Ma ch 2025.
[9] John Goodhue, Julie Ma, Ad ian Del Maes o, Sia Naja i, B uce Segee, Sco Valcou , and Ralph Zo ola.
No heas Cybe eam P og am–A Wo k o ce De elopmen S a egy o Resea ch Compu ing. The Jou nal o
Compu a ional Science Educa ion, 11(1):8–11, Janua y 2020.
[10] Thomas Hacke , P es on Smi h, Dana B unson, Lisa A a une, Thomas Chea ham, and Ewa Deelman. Building
he Resea ch Inno a ion Wo k o ce: Challenges and Recommenda ions om a Vi ual Wo kshop o Ad ance
he Resea ch Compu ing Communi y. In P ac ice and Expe ience in Ad anced Resea ch Compu ing, pages 1–7,
Bos on MA USA, July 2022. ACM.
[11] Pa ick Schmi z, Sco Yockel, Clai e Mizumo o, Thomas Chea ham, and Dana B unson. Ad ancing he Wo k-
o ce Tha Suppo s Compu a ionally and Da a In ensi e Resea ch. Compu ing in Science Enginee ing, 23(5):19–
27, Sep embe 2021. Con e ence Name: Compu ing in Science Enginee ing.
[12] Anselm S auss and Julie M. Co bin. Basics o quali a i e esea ch: G ounded heo y p ocedu es and echniques.
Basics o quali a i e esea ch: G ounded heo y p ocedu es and echniques. Sage Publica ions, Inc, Thousand
Oaks, CA, US, 1990. Pages: 270.
5
