Team 85_8H_AI corporate learner platform

To make corporate onboarding and training more effective, we use generative AI to aid employees exploit the benefits of complementary study techniques curated for each learning material.

Comment

Inspiration

Learning can be made more effective with applying the best studying technique for each type of learning material. Besides students, learning takes place at the corporate level frequently too -- during onboarding, training and certification. High-skilled sectors (e.g. Security, Aviation, Engineering) require employees to maintain currency by sitting for tests and exams. To make learning and revision more effective, thereby maximising employees' time spent doing productive work, we thought -- why not incorporate our experiences as students to fuse research-backed studying techniques with the corporate life? And why not have generative AI aid us in this goal?

What it does

An application where employees and employers can:

  • Upload Learning Material: Users can upload their learning material (e.g., course notes, lecture slides, past test questions).
  • Generative AI Detection: The platform uses generative AI to detect the type of content.
  • Matching Study Techniques: Based on the detected content type, the platform matches the best study technique.
  • Content Generation: It generates either flashcards (for memory-heavy content) or active-recall quizzes (for application-heavy content) for employees to train themselves.

How we built it

Using ReactJS as the Frontend and using Tailwindcss, with NodeJs + ExpressJs for the backend, we also integrated Supabase which uses PostgresSQL and comes with their own Authentication

Breakdown: Upload PDF -> Blob/Storage -> Get URL -> Send URL to Azure Document Intelligence -> Receive Text -> Send Text to ChatGPT to format to {Front:"", Back:""} -> Save Flashcard to database with Course ID -> Send the Text received from document Intelligence to chatGPT to generate quiz -> Save to Quiz with Questions with Question Choices to Database with Course ID

Challenges we ran into

  1. Time limit: We realised our project could not have been as feature-packed as we would have hoped for it to be, and was able to only get the flash-card feature out in time
  2. Theme was too broad: We were a little confused at the start due to the broad theme as we were not able to set a proper direction to where we should take the project
  3. Relevance of project: We initially went with the idea of AI-generated flash cards to students to improve learning, but we realised the scope was too narrow and there was not much selling point. To expand our horizons, we brainstormed other ideas and managed to fuse improving experiences at the workplace with learning -- hence our project scope was targetted towards corporate learning. We even found out that there was a huge market for corporate learning solutions, and knew we were on the right track to improve the current way things are done in the industry.

Accomplishments that we're proud of

  1. Delegating work effectively
  2. Managing stress and uncertainty effectively
  3. Clear and direct communication with one another

What we learned

  1. Fusion of ideas during the brainstorming segment goes a long way
  2. Learning can be found everywhere -- not just in schools, and there will always be a market for it
  3. Generative AI can speed up the way we learn, allowing us to be more productive in our everyday lives

What's next for Team 85_AI corporate learner platform

  1. Feature Expansion: Expand features to exploit a wider range of studying techniques (e.g. spaced repetition, mind-mapping, pomodoro) to provide corporate users with a full suite of personalised and enhanced learning experience.
  2. Dashboard Incorporation: Incorporate dashboards to view employee learning analytics, providing HR teams to better understand the difficulties or opportunities with employee training
  3. Gamification: Gamify learning experience with data from analytics, boosting morale and promote healthy competition within the organisation
Share this project:

Updates