Smaïn Fet­tem, who grad­u­at­ed from the UTC in 2023 major­ing in bio­log­i­cal engi­neer­ing, is cur­rent­ly a research engi­neer at the Euro­pean Genom­ic Insti­tute for Dia­betes (EGID), an inter­na­tion­al insti­tute ded­i­cat­ed to research into dia­betes (types 1 and 2). He is work­ing on AI tools designed to pre­dict whether a per­son is like­ly to devel­op type 2 dia­betes. Por­trait of a young man who, despite run­ning into a few pot­holes in the road, was always able to bounce back.

The first set­back came in 2016, after he had passed his sci­ence bac­calau­re­ate, when he enrolled in med­ical school. ‘At the time, treat­ing peo­ple seemed like a won­der­ful mis­sion’, he con­fides. How­ev­er, he soon realised that this was not the path he want­ed to fol­low. Think­ing back to his child­hood dream of becom­ing an inven­tor to help oth­ers, he decid­ed after two months to join a “prepara­to­ry class” at his lycée. A few months lat­er, he left the prepara­to­ry class, his sec­ond set­back, but this gap year enabled him to clar­i­fy his edu­ca­tion­al and pro­fes­sion­al ambi­tions. ‘I realised that what I want­ed was to work in bio­med­ical engi­neer­ing, a field that devel­ops devices to improve patient care’, he explains.

Seen in this per­spec­tive, rhz uni­ver­si­ty UTC, renowned for its excel­lence in bio­med­ical engi­neer­ing, was the obvi­ous choice for Smaïn Fet­tem. How­ev­er — and this was to be the final pit­fall — after his appli­ca­tion was assessed, he was refused admis­sion. He did­n’t give up, how­ev­er and opt­ed for the IUT A in Lille, where he began stud­ies in elec­tri­cal engi­neer­ing and indus­tri­al com­put­ing (GEII), see­ing it as a pos­si­ble gate­way to what was his pas­sion: bio­med­i­cine. For two years, he gave him­self the means to achieve his ambi­tions and fin­ished among the top-rank­ing stu­dents in his year. With his Lille degree, he returned to the UTC, this time being suc­cess­ful­ly admit­ted in the bio­log­i­cal engi­neer­ing pro­gramme. ‘This time, I was on the right track. After a first year in a rather gen­er­al branch, I was lucky enough, at the begin­ning of the sec­ond year, to do a place­ment at the Sig­nal and Image Pro­cess­ing Lab­o­ra­to­ry (LTSI) in Rennes. It was there that I began to work on AI and to grasp its full poten­tial in a wide range of fields, par­tic­u­lar­ly bio­med­ical domains’, explains Smaïn Fet­tem. His end­of- stud­ies place­ment took him to Car­mat, a com­pa­ny spe­cial­is­ing in arti­fi­cial hearts. There, he took a break from AI and worked main­ly on sig­nal processing.

How­ev­er, he soon got back into AI, since, with his UTC diplo­ma in hand, he joined EGID as a research engi­neer. ‘We’re work­ing on tools to pre­dict whether or not a per­son will devel­op type 2 dia­betes dur­ing their life­time. In con­crete terms, we can say that the DNA between indi­vid­u­als is 99.9% iden­ti­cal, with the remain­ing 0.1% con­sti­tutes genet­ic vari­a­tions between indi­vid­u­als, or more com­mon­ly known as ‘vari­ants’. We have a data­base of sev­er­al mil­lion vari­ants, and that’s where the role of AI is impor­tant, more specif­i­cal­ly the large lan­guage mod­els capa­ble of pro­vid­ing huge amounts of input data, which will enable us to pre­dict whether a giv­en per­son will devel­op dia­betes, and which par­tic­u­lar vari­ants or com­bi­na­tions of vari­ants will be respon­si­ble for it’, he says. $

His expe­ri­ence at UTC? ‘I par­tic­u­lar­ly appre­ci­ate the fact that we are treat­ed as future engi­neers and not as stu­dents. Hence the impor­tance giv­en with­in this uni­ver­si­ty to the devel­op­ment of our crit­i­cal sense, but also to the acqui­si­tion of real auton­o­my and a spir­it of inno­va­tion, but not ignor­ing the ethics and impact of that same inno­va­tion. Engi­neers need to take a step back and realise that they have a duty of respon­si­bil­i­ty with regard to the tools they are going to devel­op. Will they help or harm peo­ple? That’s the key ques­tion’, he concludes.

BIO EXPRESS

• 2020: admis­sion to UTC
• 2021: place­ment in the LTSI-Rennes (Sig­nal and Image Pro­cess­ing Lab­o­ra­to­ry) and his first encounter with AI
• 2024: took up research inten­sive engi­neer­ing post at the Euro­pean Genom­ic Insti­tute for Diabetes

MSD