Engineered Living Materials (ELMs) are rapidly emerging providing properties highly sought-after in material science, such as programmability, self-healing and energy-autonomy. For ELM synthesis, genetic programs written by researchers are executed by living matter, ultimately yielding the desired mechanical, structural, and functional properties of the material. However, the development of ELMs is strongly limited by a lack of understanding of how specific genetic programs translate into mechanical material properties, making ELM development a time consuming and rather artisanal process. To address this challenge, here we present a data-driven approach that enables the prediction of ELM properties based on their genetic information. We focus on mechanical properties, specifically the storage modulus (G’), since it is crucial for various real-world applications in biomedicine, construction and wearable materials. First, we engineered a collection of ELM hydrogels with tunable mechanical properties by fusing different repeats of genetically encoded PAS polypeptides, which emulate the biophysical properties of polyethylene glycol (PEG), to CsgA, the main structural unit of curli, a protein-based nanofiber produced by bacteria. We observed that incorporation of PAS repeats reduced the ELM G’, with longer repeats correlating with lower G’ values. Using data gathered from the ELM library, we trained and cross-validated a predictive model that correlates the genetic parameters, such as the presence or absence or CsgA and the number of PAS repeats fused to CsgA, with the storage modulus.  This model enabled the prediction of genetic configurations required to achieve a desired stiffness. We anticipate that our data-driven approach will be transferable to predict additional material properties based on their genetic information. These findings highlight the potential of data-driven approaches to accelerate the design and development of ELMs with tailored properties.