A large private university · GPT-4o + RAG counselling · Azure Document AI · Video Interview Analysis
Prospective students waited around 8 hours for answers, limiting personalization and 24/7 support across geographies.
Application documents required manual checking and scoring, with roughly 15 minutes of review work per applicant.
Interview evaluation queues could take about 14 days, slowing admissions decisions and candidate communication.
Fragmented admissions stages reduced efficiency, slowed the candidate pipeline, and weakened student experience.
GPT-4o + RAG answers student queries, recommends programs, and returns guided responses in about 15 seconds.
Azure Document AI extracts data, verifies documents, detects fraud signals, and completes automated checks in under 2 seconds.
Speech and OpenAI analytics transcribe responses, evaluate content, and generate faculty-ready scorecards within 1 day.
FastAPI services, CI checks, security scanning, load testing, and analytics reporting made the platform operationally supportable.
Counselling response time reduced from hours to seconds.
Document processing moved from manual review to near-instant automated checks.
Interview turnaround compressed from a two-week backlog to next-day review.
Modeled enrollment lift, with roughly $0.4M Year 1 upside and an 18-month ROI path.