Ellucian Journey — AI-Powered Career Pathways Platform by Abdul GhafoorEllucian Journey — AI-Powered Career Pathways Platform by Abdul Ghafoor

Ellucian Journey — AI-Powered Career Pathways Platform

Abdul Ghafoor

Abdul Ghafoor

Project Context

Ellucian Journey is a multi-tenant SaaS student success platform designed to bridge the gap between academic progress and workforce readiness. Students can explore career paths, visualise their skill progression, and receive personalised course recommendations — all surfaced through a unified interface embedded in the Ellucian Experience portal.
The platform ingests live academic data from institutional ERPs (Banner, Colleague) via Ellucian Ethos change notifications, keeping student records, registrations, grades, and course sections continuously synchronised across the system.

Challenges

Multi-tenancy at scale: Every data operation is scoped by tenantId from the request context down to the database query, requiring rigorous architectural discipline across six separate microservices to prevent data leakage between institutions.
AI recommendation engine: Building a meaningful recommendation system with sparse student data required integrating AWS Bedrock (Titan embeddings) with MongoDB vector search. When a student had no career path or registration history, the system fell back to popularity-based recommendations; as data accumulated, it shifted to cosine similarity over stored vector embeddings — all without a dedicated vector database, keeping the stack lean.
Custom skill map visualisation: The "constellation" skill map could not be built with any off-the-shelf charting library. It required raw D3 primitives (d3-selection, d3-zoom, d3-drag, d3-hierarchy) to implement a hexagonal orbit grid where inner rings represent a student's learned skills and outer rings represent career path targets — with zoom, drag, pan, and accessibility mode all custom-built.
Event-driven ERP ingestion: Consuming Ethos change notifications for academic levels, sections, grades, instructors, and student records in real time — with batch fallback for historical data — required careful sequencing to avoid duplicate records across high-volume change streams.

Outcomes

Delivered a production SaaS platform deployed across multiple Ellucian customer institutions, providing students with AI-driven career path and course recommendations.
Replaced manual skill tracking with an interactive visual skill map that contextualises a student's current abilities against their chosen career trajectory.
Built on Ellucian's platform infrastructure serving 3,000+ institutions globally, with the recommendation and visualisation features designed to scale horizontally per tenant.
Like this project

Posted Oct 20, 2025

Full-stack senior engineer with end-to-end ownership: AI recommendation engine, custom D3 visualizations, 6 Node.js microservices, and AWS infrastructure for a multi-tenant EdTech SaaS platform.