F om Insigh o Impac : Building
E hical, Agile, and Human-Cen e ed
AI Sys ems
Sa a GOZE | AI & Digi al Heal h Inno a ion Leade | Founde , Cogni ek Digi al | IEEE Senio Membe
In i ed Keyno e – Fo um o Ad anced Technologies (TechMomen um 2025, Se bia)
June 12, 2025
Hello om Tu kiye!
Fi s ime in Se bia
I al eady wish I had come soone — hank you o he wa m
welcome
Le 's di e in
Ready o explo e AI inno a ion oge he
Why This Talk Ma e s
Women in AI: Jus 22%
Only 22% o he global AI wo k o ce
a e women (WEF, 2023), c ea ing a
c i ical gap in pe spec i e and
inno a ion.
Tech Leade s: 1 in 10
Women hold only 1/10 o leade ship
posi ions in echnology companies
(BCG, 2022), limi ing di e se decision-
making.
Bias T ans e : 100%
Biased da a c ea es biased AI sys ems
a an ala ming a e, a ec ing
p oduc s, se ices, and decisions
wo ldwide.
"I ech is he u u e, who builds i ma e s."
My Unlikely Pa h in o Tech
F om medicine o ma hema ics, ac oss indus ies o AI, my non adi ional jou ney combined di e se expe iences wi h
de e mina ion.
Failed med school
d eam → Pu sued
Applied Ma h, Da a
Science
Pi o ed om medicine o
ma hema ics and da a
analysis.
Wo ked in Oil & Gas,
Telecom, Heal hca e,
Supply Chain
Buil c oss-sec o expe ise
ha shaped my unique
pe spec i e on echnology.
Comple ed Mas e 's
deg ee while aising 2
kids, coding a nigh ,
and wo king ull- ime
Balanced amily, educa ion,
and ca ee h ough
esilience and
de e mina ion.
Buil my i s AI model
Applied my di e se
backg ound o c ea e AI
solu ions b idging indus y
knowledge gaps.
Wha B eaks Digi al P ojec s?
Con used oles
Da a Scien is ≠ Da a Analys ≠ Da a Enginee ≠ BI de elope
Buzzwo ds > cla i y
Fancy e ms hiding simple concep s
No sha ed language
Tech & business disconnec
"The opposi e o simplici y isn' complexi y — i 's con usion." – Edwa d Tu e
Case S udy – Medical Supply Chain
Challenge
Inconsis en 3PL da a ac oss 6 egions globally
Ac ion
Buil SAP/non-SAP ETL, ha monized KPIs
Resul
-40% manual wo kload, 3x as e epo s
"Less is mo e." – Mies an de Rohe
* No b and names shown due o con iden iali y.
Case S udy – AI Assis an o
Diabe es
Pe sonal Mo i a ion
Family membe wi h Type 1 Diabe es
Despi e ad anced echnology, dosing
decisions emain complex and
s ess ul
AI Solu ion
In elligen assis an using con inuous
glucose moni o ing da a
Combines insulin sensi i i y, nu i ion,
and s ess ac o s o comp ehensi e
suppo
Human-Cen e ed Design
Ne e o e ides human judgmen
Suppo s decision-making while p ese ing pa ien au onomy
"Fi s , do no ha m." – Hippoc a es
Case S udy – COVID Consume
Beha io Analysis
Analyzed Mobili y
Tu key's mo emen pa e ns du ing lockdown
Key Finding
G oce y + pa k ac i i y su ged unexpec edly
Insigh
Habi s change slowe han echnology
Lesson
Build sus ainable ech, no hype
Wha Real Agili y Looks Like
Empowe eams
Gi e au ho i y wi h esponsibili y
Build eliable wo kspace and employee loyal y
C ea e space whe e ailu e is lea ning
P io i ize explainabili y
Unde s anding umps echnical pe ec ion
"Fi s help. Then ask."
