All Posts
Search, Data & Cloud
EdTech

Intelligent Educational Program Matching for Student Success

A student card and a results list of programs, each with a chance ring, the top ring filled in red.
Published on
November 20, 2024

Problem

Users faced challenges in finding educational programs ranked by the likelihood of successful admission after applying. The search experience was not personalized based on user profiles, resulting in a lack of tailored recommendations. Additionally, the system exhibited performance issues in specific scenarios, and the behavior of the full-text search functionality was unclear to users, reducing confidence in the results.

Also applicable to

This problem is common in industries relying on personalized search experiences, such as e-commerce platforms recommending products, job portals tailoring job matches, and healthcare systems assisting patients in finding suitable treatments or providers.

Solution

We designed and implemented a modern, intelligent search system leveraging historical application data. A machine learning model was developed using AWS SageMaker and PyTorch to calculate a personalized success score for each educational program for individual users. The data was indexed using AWS OpenSearch, enabling faster, more relevant search results. The new system also introduced enhanced filtering and ranking capabilities tailored to user needs.

Throughout the project, our team provided architectural guidance, oversaw the design and implementation phases, and managed the deployment of this scalable, high-impact solution.

Impact

The enhanced search system empowered users with faster, more relevant results and a clearer understanding of their likelihood of success for each program. This resulted in an improved user experience and increased confidence in the search tool’s reliability and relevance.

Technologies

  • AWS OpenSearch: High-performance search and data indexing.
  • AWS SageMaker: Model training and deployment for calculating personalized success scores.
  • PyTorch: Developing and refining machine learning models.

Our expertise in machine learning, data science, and cloud-based technologies ensures solutions that deliver measurable impact for businesses. With a focus on ROI and scalable results, we provide end-to-end support, from discovery to implementation.

Soroush Mehri
Written bySoroush MehriCo-Founder and Chief AI Officer

Ex-Microsoft. 10+ years of experience in AI/ML, MSc from Université de Montréal (Mila). US patent holder.

Mojtaba Ebadi
Reviewed byMojtaba EbadiCo-Founder and Chief Operating Officer

Sr. Engineering Leader with 15+ years of experience. Ex-ApplyBoard. AWS Certified Solutions Architect.