This research investigates the impact of microwave power, processing time, and solid-to-solvent ratio on protein recovery from foxtail millet (Setaria italica), using an artificial neural network (ANN) and genetic algorithm (GA). The extracted protein and subsequent hydrolysates were also evaluated for their techno-functional, structural, and digestibility properties. The ANN model, trained with the Levenberg-Marquardt algorithm and optimized by a GA, identified optimal extraction conditions (960 W, 66.14 s, 0.1264 g/mL), achieving a protein recovery yield of 30.02 ± 0.97 %. Protein recovery increased 1.15-fold for raw microwave-treated protein and 1.52-fold for germinated microwave-treated protein. The functional properties, including foaming, emulsifying, and digestibility were improved. Germinated microwave-treated protein hydrolysates (GMPH) exhibited the highest soluble protein content (17.79 ± 0.33 mg/mL) and degree of hydrolysis (7.63 ± 0.25 %) at 2.5 % (v/w) enzyme concentration, while raw microwave-treated protein hydrolysates (RMPH) showed 16.59 ± 0.36 mg/mL and 5.57 ± 0.14 %, respectively. Partially purified peptides from GMPH and RMPH showed strong antioxidant and angiotensin-converting enzyme (ACE) inhibitory activities, with GMPH showing the highest bioactivity. These findings highlight the potential of germination and microwave-assisted processing to produce high-quality protein and bioactive peptides, supporting the development of protein-rich foods and nutraceuticals to address global protein demand.