Abstract: This pilot randomized controlled trial evaluated the efficacy of a veteran peer-mentoring intervention designed to enhance academic performance in the MachinE Learning for VETerans (MELT-VET) micro-credential program. Twenty-seven U.S. military veterans were randomized to either a peer mentoring group (n = 10) or a control group (n = 17). Peer mentors met with participants to review academic progress and address psychosocial goals using a contract-based framework. Academic performance was assessed via pretest and posttest scores on a ten-item quiz measuring proficiency in machine learning coding. Linear mixed-effects models, adjusted for covariate differences at baseline, estimated changes in scores over time. Groups were comparable at baseline. The peer-mentored group improved 38.8% from pretest (M = 6.7, SE = 0.4) to posttest (M = 9.3, SE = 0.2), while the control group improved 2.4% (M = 8.3, SE = 0.4 to M = 8.5, SE = 0.5). A significant time-by-group interaction (b = 1.0, SE = 0.3, p < 0.01) indicated greater gains among peer-mentored participants in an adjusted model. Structured peer mentoring significantly enhanced veterans' learning outcomes in a technical training context, suggesting that formalized mentoring frameworks may strengthen academic adjustment during the transition to civilian education.