Built core backend infrastructure and AI integration layer for Aiberry — an anonymous mental health screening platform used by corrections facilities, universities, corporate wellness programs, and behavioral health organizations nationwide.

Developed the real-time AI analysis pipeline that processes user responses through LLM-powered assessment models, generates structured mental health insights, and delivers actionable screening results to clinicians and administrators. Built multi-tenant architecture with strict role-based access control — patients, clinicians, researchers, and admins all accessing shared data with isolated permissions and HIPAA-compliant data handling.

Engineered the full stack: FastAPI backend, React frontend with interactive screening interfaces, PostgreSQL database with row-level security, Redis caching for low-latency session management, and AWS deployment with observability and audit logging. Integrated AI models for natural language understanding and behavioral signal analysis, with confidence scoring that flags uncertain assessments for clinical review rather than presenting them as definitive. The platform serves as a digital front door to mental health — reducing barriers to screening and connecting people with care faster than traditional intake processes.

Tech: Python, FastAPI, React, TypeScript, PostgreSQL, Redis, AWS, OpenAI API, Docker