1. General

Category

Category

Education

2. Project Details

Company or Institution

uPlanner

Project

Student Succes

General description of the AI solution

As a company, uPlanner provides a wide range of data-driven solutions for Higher Education Institutions (HEI). Algorithms and data are at the core of our solutions, promoting efficient academic planning, resource management, curriculum improvement, and student retention. In the context of this proposal, we present and describe two solutions for student success: uRetention (web platform) and uExperience (mobile app), with a direct impact on the fourth SDGs (Education Quality). uRetention and uExperience help institutions detecting which students require help and enhancing the student engagement and communication with peers and faculty. More than 300,000 students are using these solutions on over 11 institutions, and we expect to double this numbers before 2023. Both solutions have Artificial Intelligence (AI) algorithms at its core, integrate AI-as-a-service when available and applicable, and develop custom data processing and algorithms when required. They have been narrowly tailored to the needs and particularities of higher education. As with all uPlanner’s portfolio, these solutions have a clear limitation of purpose, where data collection and processing have been created to solve a specific educational problem, respecting the applicable restrictions on data legislation. Higher Education Institutions (HEI) in Latin America are very demanding due to the ample differences among institutions in terms of complexity, size and technological infrastructure. The knowledge acquired on over 100 implementations carried out on Latin America gives us confidence on the potential and applicability of our solutions. The scaling of these solutions is guaranteed as they have been built as Solution-as-a-Software (SaaS) on top of Azure Services by Microsoft, as we continue to learn, grow and evolve them to reflect the needs of HEIs.

Website

https://uplanner.com/en/

Organisation

uPlanner

3. Aspects

Excellence and Scientific Quality: Please detail the improvements made by the nominee or the nominees’ team or yourself if your applying for the award, and why they have been a success.

We have developed and implemented two solutions to solve specific aspects of Student Success. The first solution is uRetention, a SaaS platform that connects to institutional systems, collects and provides a risk assessment for each student. The second solution is uExperience, a mobile app that gives access to all digital services available at the institution, delivering answers through a chatbot, serving as messaging app and more. In all cases, the algorithmic approach has been to leverage existing tools when available and build upon them to create solutions specifically tailored to the needs and particularities of Higher Education Institutions (HEI). With this approach, uExperience uses existent AI-as-a-service tools for the chatbot and intelligent event suggestions, based on text analysis. This allows it to use state-of-the-art, stability and cost-effective results in terms of Natural Language Processing. uRetention, by contrast, uses an interpretable machine learning algorithm specifically designed to predict HEI dropouts, allowing the institution to provide pre-computed dropout factors, evaluate data quality, and configure thresholds, and create predictive models for specific student segments. We highlight these characteristics as we have learned that institutions can be highly diverse: in size and complexity, types of careers and students (regular/weekend/online), legal status (public/private), and mission focus (teaching/research). Solutions in HEI must work successfully despite these huge differences. While the uRetention and uExperience solutions are currently being used by clients in Latin America, we envision the need of a third solution, currently under development, to provide a broader and unified platform of Student Success, where data integration, data collection and data consolidation work simultaneously to guarantee that all needs regarding Student Success are covered. Indeed, efficient management and matching available success interventions (nudges) to students is still largely unresolved problem.

Scaling of impact to SDGs: Please detail how many citizens/communities and/or researchers/businesses this has had or can have a positive impact on, including particular groups where applicable and to what extent.

uPlanner’s solutions directly impact the goal of Quality Education SDG, particularly targets 3 (equal success), 4 (relevant skills), and 5 (gender equality and inclusion). Reducing dropouts and creating programs that are aligned with the labor market addresses several SDG’s (5, 8, 10). As a company, we have been working with 100 institutions in 16 countries, impacting the lives of nearly 2,5 million students. We have developed our solutions to be Software-as-a-Service (SaaS), to be able to scale up fast and efficiently, with the internationalization and personalization required for world-wide usage. We hope to continue to grow at an even faster rate to accomplish our mission of helping institutions adopt a data-driven approach. Our uRetention solution provide HEIs with a tool to prevent dropouts. A predictive model is built with widely available information such as academic progress, grades, and socioeconomic background, which also allows to include new risk factors as well as new functionalities that can adapt to a variety of local needs and realities across countries. Similarly, uExperience allows to improve student engagement, providing a simplified experience to digital services, help and communication. In doing so, our solutions impact on other SDG’s such as No Poverty and Reduced Inequalities, as education stands as one of the major drivers of socioeconomic mobility, while disparities in access and graduation rates tends to affect minorities and disadvantaged populations adversely. In Latin America and other third world countries, smarthphones have become more frequent than personal computers so providing access to education through mobile app is a key issue. An internal tool allows us to obtain quantitative and qualitative impact assessment of our solutions. We already have indicators to measure satisfaction with our uExperience app trough NPS methodology, alongside with successful case studies of clients with positive experience using uRetention and uExperience.

Scaling of AI solution: Please detail what proof of concept or implementations can you show now in terms of its efficacy and how the solution can be scaled to provide a global impact ad how realistic that scaling is.

uRetention and uExperience have been implemented at 11 institutions, with a combined reach of over 300,000 students. As the Covid-19 crisis impacted an educational sector that was already highly stressed, we expect solutions that help addressing student success to play a central role in reshaping the relationship between students and institutions. In terms of impact, the “uExperience app” developed in partnership with and for our client Universidad de los Andes in Colombia (the country’s foremost private university), is the top valued app for an institutional application on the AppStore, with similar positives reviews on the Google Play Store (Android Market). Similarly, the Machine Learning algorithms of uRetention are constantly being tested through the different datasets implemented by our clients. A particularly interesting result was obtained in a collaboration experiment with Universidad Nacional de México, where the algorithm trained with data from 6 years of high school data, obtaining results with an accuracy over 95% and an f1-score over 75%.

All our solutions are built using Microsoft’s Azure Platform, therefore being able to scale without restrictions. A key component of our solutions is obtained through the communities we serve. We constantly search for the valuable feedback on usability and needs, and we iterate quickly over new ideas. This fast discovery and delivery cycles allows to respond and adapt to what is required by Higher Education Institutions. uPlanner solutions integrate seamlessly with the ecosystem of other educational tools, such as Student Information Systems (SIS), Learning Management Systems (LMS), and even Customer Relationship Management (CRM) and Enterprise Resource Planning (ERP). In this way, we establish long-term relationship with each institution, and constantly search to form partnerships with other EdTech companies and startups.

Ethical aspect: Please detail the way the solution addresses any of the main ethical aspects, including trustworthiness, bias, gender issues, etc.

All uPlanner solutions have a clear and straightforward purpose for collecting and processing data, known and approved by the institution. We have worked with external consultants in order to make sure that our solutions respect applicable laws and regulations, making uPlanner’s solutions GDPR compliant.

Regarding uExperience, there has been a deliberated decision to not use more information that strictly necessary. It does not display adds nor share student information to other parties, nor require more permissions than strictly needed for the correct display and performance of the application. All messages on the app travel in an encrypted format, so that only sender and receiver can read them. The app chatbot allows to provide large scale student assistance, and despite been monitored for improvement, data is anonymous and aggregated, so individual student information regarding questions or answers cannot be identified.

With respect to uRetention, the machine learning algorithm has in-place subroutines to detect missing, wrong, or out of range data, so the model training process can be trustful. As part of the implementation phase, the model is tested on a finished semester to check the prediction, gain knowledge on the quality of the model and to correct any mistake. The solution was conceived to complement the work of the Student Success team, not to replace them. It provides alerts and clean, reliable information and dropout risk assessments, but does not automatize the decisions and makes sure to keep the “human in the loop”. A student that has been deemed as high risk by a human but wrongfully marked as low risk by the algorithm, can be manually reclassified and taken into consideration. The solution is not a “gatekeeper”; is not used to select candidates at the admission stage, but to assist institution detect more easily which students require help.