Sliders is an intelligent, web-based platform designed to redefine how students engage with recorded lectures. By combining speech recognition with semantic analysis, Sliders automatically maps lecture audio to corresponding slides, diagrams, and content blocks—giving users a precise, click-to-audio experience. Ever spent hours scrubbing through long recordings to find “that one explanation”, with Sliders users can upload a lecture video and slide deck, and within minutes, receive an interactive playback environment featuring searchable transcripts, synced slide views, and targeted audio snippets.
Name: Gagandeep Singh
Student number: 48268503
The system will provide users with a more interactive lecture experience. Each text block/diagram/image will be attached with its corresponding audio snippet from the lecture. Each identified piece of information could then be connected through a mind map.
1. User Login / Signup
Users begin by creating an account or logging in to an existing one. Each user has a personal dashboard where they can upload new lecture content, access past uploads, join study groups, and view shared resources.
2. Upload Portal
Users can upload a lecture video and accompanying slides(PDF or PowerPoint. Uploads made outside a study group are private to the user unless shared explicitly. The backend immediately starts processing the uploaded files.
3. Study Groups (Institutional Sharing)
Universities and schools can create verified Study Groups that users can join. When a lecture and slide deck are uploaded into a group, they are automatically processed and shared with all members—so only one user (typically an instructor or class rep) needs to upload the content. Group members can then access and interact with the material independently.
4. Audio-Slide Alignment
Sliders uses speech recognition and semantic analysis to transcribe the lecture and automatically segment and align the audio with the relevant slides, diagrams, or text blocks. This mapping ensures users can easily navigate to the specific parts of the lecture they need to review.
5. Interactive Playback Interface
After processing, the user is presented with a playback environment where they can view slides and click on any slide or section within the slide to play the exact portion of the lecture audio that matches it reducing time spent searching manually through video timelines.
6. Full Transcript Generation & Viewer
Sliders provides a full transcript of the lecture, shown alongside the slide viewer. The transcript is searchable, scrollable, and interactive—clicking any line will trigger the related audio and highlight the matching slide.
7. Concept Board – Visual Knowledge Builder
Users can extract identified elements (such as images, diagrams, or text blocks) from any processed lecture and drag them onto a collaborative visual workspace, similar to Figma or Miro. On this Concept Board, users can rearrange content freely, draw connections using arrows, labels, and notes, and build mind maps or custom visual study sheets from multiple lectures. This tool helps students connect concepts across different topics and enhances long-term retention.
The scope for the MVP of this system will include the following features:
1. User Authentication and Dashboard
Users will be able to create an account, log in, and access a dashboard listing their uploaded content and any joined study groups. Basic session management and account persistence will be implemented.
2. Upload Portal (Slides + Video)
Users can upload a lecture video and its corresponding slides (PDF or PowerPoint). These will be stored and passed to the backend.
3. Audio Transcription and Slide Alignment
The backend will extract audio, generate a transcript using a speech-to-text engine (e.g., Whisper API), and segment the transcript. These segments will then be aligned with slide content based on timing and semantic similarity. Each slide (or its key content elements) will be linked with an audio snippet.
4. Interactive Playback Interface
Users can view processed slides and click on them to trigger playback of the associated audio. The transcript will be displayed alongside the slides and kept in sync during playback. Clicking on the transcript will also jump to the relevant part of the lecture.
5. Study Group Access (Basic)
Users can join existing study groups. Content uploaded to a group will be shared across all group members. Group creation and invitations may be handled manually for this stage.
Scalability
Although the MVP is designed for a small group of users, Sliders is intended to support institution-wide adoption over time. The backend architecture will allow horizontal scaling of compute-intensive components like transcription, semantic alignment, and playback indexing. Scalability is especially critical around assessment periods when thousands of users might upload or access lectures concurrently. The system must ensure low latency and prevent queue bottlenecks during high-traffic usage.
Extensibility
The MVP does not add much value to the current learning experience but lays important fundamentals for future capabilities with actual innovation —such as interactive concept boards, automatic summarisation, ECHO360/LMS integration, support for collaborative annotation, and alternative input types like YouTube videos or live Zoom recordings— which should be seamlessly incorporated.
Availability
Sliders must be accessible to users at any time. ‘On Demand’ in this case would mean that the platform should provide responsive interactions, quick load times, and consistent uptime, even under varying network conditions or device types. Users should also be able to download the annotated content. Availability also includes ensuring that processed lecture content remains accessible after uploading without unnecessary delays. Cross-platform compatibility—including desktop and mobile browsers—will be supported to allow flexible, on-the-go usage.
Availability
Uptime will be monitored during development and testing using a simple ping or health-check.
Manual and automated tests will validate responsiveness of core
features, including upload, playback, and transcript interaction.
The system will be tested across different browsers and devices to ensure consistent performance.
Scalability
Stress tests will simulate multiple users uploading and accessing content concurrently to evaluate backend performance.
Response times for audio transcription and playback loading will be logged at different simulated load levels.
Tests will verify the system’s ability to run multiple transcription tasks in parallel.
Extensibility