The use and misuse of antibiotics promotes the spread of antibiotic resistance. This issue is of particular importance in urinary tract infections (UTIs), which are among the most common bacterial infections worldwide. While the pathogens causing UTIs are commonly resistant to different antibiotics, treatment is often chosen empirically, in the absence of antibiotic susceptibility testing, risking ineffective treatments and adverse outcomes. Using a combined computational-experimental approach we are currently developing algorithms which, based on demographics and clinical data, predict an infection-specific profile of resistance, and suggest an optimized personally tailored treatment accordingly. These algorithms have the potential to reduce mismatched antibiotic treatment, both improving healthcare and helping in the global effort to impede the antibiotic resistance epidemic.