Lectrify.app

Lectrify empowers a presenter, and energizes an audience. It dynamically creates a concurrent quiz experience upon any question asked towards the audience. Get answers in real time.

Comment

Introduction

“Large-lectures allow students to be passive, anonymous and unengaged”

One of the biggest problems any presenter faces is engagement, especially when asking a question to a large audience. Often met with silence and no show of hands, this experience is frustrating; even if one person answers, it’s hard to get a grasp on how well the audience actually fares with the question.

What if all the lecturer had to do was ask the question, and it pops right into each person’s hands?

This is what Lectrify solves. A lecturer asks a relevant question with just their voice, and it sends a multiple choice question to everyone connected in the audience, providing the host instant insight.

It offers a concurrent quiz experience, where someone doesn’t have to stop presenting to host a quiz that runs alongside the show. It integrates seamlessly, runs concurrently, and updates dynamically to be as out of the way as possible.

Without a sense of competition built into the app, it is likelier to display honest answers, aiding the invaluable understanding of how well an audience actually understands a topic.

Use Cases

A lecturer at UNSW could use Lectrify to gain instant responses from their students without having to stop the lecture. A workplace admin could use Lectrify to make sure people understand policy A presentation happening across online and in-person could be better connected with Lectrify

Integration into UNSW

Lecturers or tutors can implement Lectrify within UNSW lectures or tutorials to gather insight into how well students have grasped certain concepts. Currently, Quizlet and Kahoot are popular options, but this requires that the lecturer stop their delivery to do a quiz. Lectrify allows them to ask questions and have a quiz without disrupting the lecture flow. Additionally, this allows them to quickly identify which concepts students need more assistance on. As a result, this enhances the understanding of certain concepts and the overall learning experience of students.

How it was built

The demo version of this application was built entirely with Javascript, HTML, CSS and Python.

The microphone was built with Python The backend was built with Express.js The front-end was built with JS, HTML, CSS and ChartJS

The whole experience leverages the Express.js framework, the OpenAI local whisper model and the OpenAI ChatGPT Beta endpoints to deliver close to fully functional features.

For real-time capabilities, our stat page uses simple API polling, whilst our client-quiz pages utilise EventSources to listen for any server data changes for their associated room.

We’re hosting our site on a Linux VPS, using Nginx as the reverse-proxy and Cloudflare as the domain registrar.

Scalability and Costs

The only costs associated with the project are:

Server and Hosting costs (Easily scalable) The OpenAI ChatGPT endpoints (built for scale)

For costs, the project is estimated to be $0.2-0.3 USD for every 750,000 words spoken according to OpenAI’s pricing model. Server costs vary significantly.

Just taking into account the OpenAI costs and assuming 24,000 words are spoken or generated during a single lecture, this would add a cost of $0.0064-0.0096 USD per lecture.

For scale, the project relies on easily scalable hosting and endpoints, more rooms will only increase server load and not any client-run code.

Future Features

To further enhance user experience, further development to accommodate for user needs are necessary.

For instance, many students have missed their lectures and struggle to get back on track with content. In the future, students can access the quizzes of past lectures, students can revise and catch up on content quickly with Lectrify.

Another major update that has been considered is to enhance the lecturers control over the questions that get sent through to students. We could implement a feature to grant lecturers the ability to upload questions into Lectrify before the lectures. During a lecture, a presenter will eventually be able to accept or decline which questions get sent through to the students with just the tap of a button on their phone.

To create a more bespoke Lectrify experience in UNSW, we could integrate the web app directly into the myUNSW calendar. That way students can see any lectures their courses have started and enter their enrolled question-room without a code by matching timetables instead.

+ 1 more
Share this project:

Updates