The FUNDET Foundation faces inefficiencies in manual processes, including student data errors (35%), loss of graduate traceability (42%), and delays in job placement (an average of 15 days), which increase administrative workload and reduce the social impact of its programs for vulnerable populations in Guayaquil. To address these issues, a web and mobile platform was developed and implemented to automate student registration and psychotechnical assessment using rule-based expert systems (92% accuracy), and course assignment, while enabling real-time bidirectional job matching. The system was designed to comply with usability standards (ISO 9241-11: effectiveness ≥90%, efficiency ≤2 min/task, SUS ≥85) and accessibility requirements (WCAG 2.1 AA: contrast ≥4.5:1, full keyboard navigation). The project applied RUP (web) and Mobile-D (mobile) development methodologies, implemented using PHP, MySQL, HTML5, CSS3, Bootstrap, jQuery, XAMPP (web platform), and Java with Android Studio (mobile app), following an MVC architecture with RESTful API integration. The platform was evaluated with n=52 users (95% confidence level, 5% margin of error). Results show a 67% increase in efficiency (1.8 vs. 5.5 min/task), a SUS score of 87.4/100, a 28% improvement in job placement (from 72 to 92 graduates per month), and a 19% increase in retention. In conclusion, the solution incorporates AI-enhanced psychotechnical testing, predictive language features, and a dropout-tracking dashboard (95% reliability), powered by RUP/Mobile-D, replicable in NGOs ≤$5K, with limitations in rural connectivity and future in machine learning, iOS/PWA for secure data recording.