Eladdin
Beyond Learning Technologies
Eladdin
Eladdin utilises patented AI technology to accurately detect the root cause of student mistakes. This, combined with a database of questions, gives teachers more time to tailor lessons to each child's pace.
​
Deliverables
Web App
Problem Statement
The Problem
Teachers have a lot of tasks to do everyday. They don't have the time to analyse each students' weaknesses and tailor worksheets to them. This causes many students to be left behind.
The Proposed Solution
As worksheets were done with paper and pen, teachers could not track the progress and weaknesses of the students easily. Eladdin is an A.I. technology that highlights important metrics to teachers. It also uses a database of questions and A.I. to allow teachers to create personalised worksheets quickly and painlessly.
Dashboard
Displays key insights to teachers at a glance.
Assignments
A database of A.I. generated worksheets for teachers to use.
A.I. Assistant
Recommends questions to assign each student.
Exploring
Gathering Information
To understand the working environment and user needs of the product, I carried out the following:​
-
Competitive analysis of 5 companies' UI, usability and product features to understand the current solutions in the market.
-
A survey with 15 MOE and tuition teachers to find out the information they were lacking.
-
3 in-depth interviews with MOE and tuition teachers to discover user needs.
​
Below, you can find the affinity map and competitor analysis of all these findings.
Problem 1: Too Many Tasks
First, I realised that the teachers had many tasks to do every day. A typical day involved the following:
Lesson Plans
Create lesson plans for upcoming weeks.
Worksheets
Marking and creating assignments for the next few days and weeks.
Monitoring
Track and see who was falling behind. Try to help them.
This meant that they did not have much time to help each individual child, even though many wanted to. It was not feasible given the deliverables and timeline they had. This made it very frustrating for them.
Problem 2: Solutions were not user friendly
The current solution is the Student Learning Space (SLS), created by the Singaporean government. This is an e-learning platform that lets teachers give work to students to do online. However, it is extremely difficult to use and many teachers complained how inefficient it was. In fact, one one stated, "I'd rather just make my worksheet in a word doc."
To set questions was extremely time-consuming and difficult, as teachers had to click through many dropdowns to get to the topic they wanted. Moreover, teachers could not customise the questions to suit their needs easily. Thus, paper worksheets created on Microsoft Word continue to be the most used solution.
Problem 3: Solutions do not leverage data sufficiently
SLS still required the manual adding of questions. There was no reduction in the time spent on creating worksheets. The database that was available depended on other teachers contributing to the platform. However, with the low user-friendliness of the site, few teachers had time to add questions. This led to a small database of questions.
​
Other platforms also do not utilise A.I. to uncover insights into students' progress. While they saved teachers time to mark worksheets, it was still impossible to analyse so many students' errors. Thus, teachers still could not target individuals with specific questions.
Research Outcomes
With the team, we prioritised the core features of Eladdin to develop first. New ideas and hypotheses were generated on what information we thought users would find most useful. With everyone aligned, it was time to create prototypes and test them out with users.
Iterating
Prototyping
To visualise and test the usability of the web app, I created these prototypes:
-
Sketches to show stakeholders and modify designs.
-
Wireframes for first 2 usability tests.
-
High-fidelity prototypes for next 3 usability tests. I used Figma.
​
Style Guide
I created a simple style created with the goal of unifying their web app and website. This included:
-
Colour palettes
-
Typography
Usability Testing and Research Outcomes
I conducted usability tests with 5 teachers. Further below I write more detailed findings, but here were the main points:
-
Teachers initially did not understand the dashboard well due to poor grouping of sections. Following Gestalt theory, I resolved this by placing related items closer together and refining which information to display.
-
At first, the flow for creating a worksheet was to first make the worksheet, then assign it to students. I realised this flow was the opposite for a teacher's mental model of assigning homework. Teachers preferred it much more when I switched the steps to first pick a class, then assign worksheets to that class.
-
Basic search was not enough for teachers looking for the right topic. With the MOE syllabus containing hundreds of topics and sub-topics, teachers found it a lot easier to filter topics using both search and an accordion filter which is suitable for such a large number of options.​
Low- to Mid-Fidelity Prototypes
Sketches, Wireframes
Research Outcomes: Series of Changes to Designs
Please click here for a summary of the major changes.
Final Design
Conclusion
Key Takeaways
Overall it was a new challenge to lead UX discussions with a team of experienced educators, administrators and technical experts. As always, usability testing proved to be an invaluable resource in this process as I learnt countless new improvements to be made on the design.
​
In Hindsight...
In hindsight, I would have incorporated more accessibility standards into more user interface design. Some of the designs would fail to meet today's standards. Also, I create a softer and less saturated colour palette for the web app which is in line with today's friendlier design.
​
Future Plans
Eladdin is currently under development, and is set to launch in mid 2021.
